Fixed MTP to work with TWRP

drivers/s390/Makefile

@@ -0,0 +1,8 @@
+#
+# Makefile for the S/390 specific device drivers
+#
+
+obj-y += cio/ block/ char/ crypto/ net/ scsi/ kvm/
+
+drivers-y += drivers/s390/built-in.o
+

drivers/s390/block/Kconfig

@@ -0,0 +1,90 @@
+comment "S/390 block device drivers"
+	depends on S390 && BLOCK
+
+config BLK_DEV_XPRAM
+	def_tristate m
+	prompt "XPRAM disk support"
+	depends on S390 && BLOCK
+	help
+	  Select this option if you want to use your expanded storage on S/390
+	  or zSeries as a disk.  This is useful as a _fast_ swap device if you
+	  want to access more than 2G of memory when running in 31 bit mode.
+	  This option is also available as a module which will be called
+	  xpram.  If unsure, say "N".
+
+config DCSSBLK
+	def_tristate m
+	prompt "DCSSBLK support"
+	depends on S390 && BLOCK
+	help
+	  Support for dcss block device
+
+config DASD
+	def_tristate y
+	prompt "Support for DASD devices"
+	depends on CCW && BLOCK
+	select IOSCHED_DEADLINE
+	help
+	  Enable this option if you want to access DASDs directly utilizing
+	  S/390s channel subsystem commands. This is necessary for running
+	  natively on a single image or an LPAR.
+
+config DASD_PROFILE
+	def_bool y
+	prompt "Profiling support for dasd devices"
+	depends on DASD
+	help
+	  Enable this option if you want to see profiling information
+          in /proc/dasd/statistics.
+
+config DASD_ECKD
+	def_tristate y
+	prompt "Support for ECKD Disks"
+	depends on DASD
+	help
+	  ECKD devices are the most commonly used devices. You should enable
+	  this option unless you are very sure to have no ECKD device.
+
+config DASD_FBA
+	def_tristate y
+	prompt "Support for FBA  Disks"
+	depends on DASD
+	help
+	  Select this option to be able to access FBA devices. It is safe to
+	  say "Y".
+
+config DASD_DIAG
+	def_tristate y
+	prompt "Support for DIAG access to Disks"
+	depends on DASD
+	help
+	  Select this option if you want to use Diagnose250 command to access
+	  Disks under VM.  If you are not running under VM or unsure what it is,
+	  say "N".
+
+config DASD_EER
+	def_bool y
+	prompt "Extended error reporting (EER)"
+	depends on DASD
+	help
+	  This driver provides a character device interface to the
+	  DASD extended error reporting. This is only needed if you want to
+	  use applications written for the EER facility.
+
+config SCM_BLOCK
+	def_tristate m
+	prompt "Support for Storage Class Memory"
+	depends on S390 && BLOCK && EADM_SCH && SCM_BUS
+	help
+	  Block device driver for Storage Class Memory (SCM). This driver
+	  provides a block device interface for each available SCM increment.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called scm_block.
+
+config SCM_BLOCK_CLUSTER_WRITE
+	def_bool y
+	prompt "SCM force cluster writes"
+	depends on SCM_BLOCK
+	help
+	  Force writes to Storage Class Memory (SCM) to be in done in clusters.

drivers/s390/block/Makefile

@@ -0,0 +1,25 @@
+#
+# S/390 block devices
+#
+
+dasd_eckd_mod-objs := dasd_eckd.o dasd_3990_erp.o dasd_alias.o
+dasd_fba_mod-objs  := dasd_fba.o
+dasd_diag_mod-objs := dasd_diag.o
+dasd_mod-objs      := dasd.o dasd_ioctl.o dasd_proc.o dasd_devmap.o \
+			dasd_genhd.o dasd_erp.o
+ifdef CONFIG_DASD_EER
+dasd_mod-objs      += dasd_eer.o
+endif
+
+obj-$(CONFIG_DASD) += dasd_mod.o
+obj-$(CONFIG_DASD_DIAG) += dasd_diag_mod.o
+obj-$(CONFIG_DASD_ECKD) += dasd_eckd_mod.o
+obj-$(CONFIG_DASD_FBA)  += dasd_fba_mod.o
+obj-$(CONFIG_BLK_DEV_XPRAM) += xpram.o
+obj-$(CONFIG_DCSSBLK) += dcssblk.o
+
+scm_block-objs := scm_drv.o scm_blk.o
+ifdef CONFIG_SCM_BLOCK_CLUSTER_WRITE
+scm_block-objs += scm_blk_cluster.o
+endif
+obj-$(CONFIG_SCM_BLOCK) += scm_block.o

drivers/s390/block/dasd_alias.c

@@ -0,0 +1,990 @@
+/*
+ * PAV alias management for the DASD ECKD discipline
+ *
+ * Copyright IBM Corp. 2007
+ * Author(s): Stefan Weinhuber <wein@de.ibm.com>
+ */
+
+#define KMSG_COMPONENT "dasd-eckd"
+
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <asm/ebcdic.h>
+#include "dasd_int.h"
+#include "dasd_eckd.h"
+
+#ifdef PRINTK_HEADER
+#undef PRINTK_HEADER
+#endif				/* PRINTK_HEADER */
+#define PRINTK_HEADER "dasd(eckd):"
+
+
+/*
+ * General concept of alias management:
+ * - PAV and DASD alias management is specific to the eckd discipline.
+ * - A device is connected to an lcu as long as the device exists.
+ *   dasd_alias_make_device_known_to_lcu will be called wenn the
+ *   device is checked by the eckd discipline and
+ *   dasd_alias_disconnect_device_from_lcu will be called
+ *   before the device is deleted.
+ * - The dasd_alias_add_device / dasd_alias_remove_device
+ *   functions mark the point when a device is 'ready for service'.
+ * - A summary unit check is a rare occasion, but it is mandatory to
+ *   support it. It requires some complex recovery actions before the
+ *   devices can be used again (see dasd_alias_handle_summary_unit_check).
+ * - dasd_alias_get_start_dev will find an alias device that can be used
+ *   instead of the base device and does some (very simple) load balancing.
+ *   This is the function that gets called for each I/O, so when improving
+ *   something, this function should get faster or better, the rest has just
+ *   to be correct.
+ */
+
+
+static void summary_unit_check_handling_work(struct work_struct *);
+static void lcu_update_work(struct work_struct *);
+static int _schedule_lcu_update(struct alias_lcu *, struct dasd_device *);
+
+static struct alias_root aliastree = {
+	.serverlist = LIST_HEAD_INIT(aliastree.serverlist),
+	.lock = __SPIN_LOCK_UNLOCKED(aliastree.lock),
+};
+
+static struct alias_server *_find_server(struct dasd_uid *uid)
+{
+	struct alias_server *pos;
+	list_for_each_entry(pos, &aliastree.serverlist, server) {
+		if (!strncmp(pos->uid.vendor, uid->vendor,
+			     sizeof(uid->vendor))
+		    && !strncmp(pos->uid.serial, uid->serial,
+				sizeof(uid->serial)))
+			return pos;
+	};
+	return NULL;
+}
+
+static struct alias_lcu *_find_lcu(struct alias_server *server,
+				   struct dasd_uid *uid)
+{
+	struct alias_lcu *pos;
+	list_for_each_entry(pos, &server->lculist, lcu) {
+		if (pos->uid.ssid == uid->ssid)
+			return pos;
+	};
+	return NULL;
+}
+
+static struct alias_pav_group *_find_group(struct alias_lcu *lcu,
+					   struct dasd_uid *uid)
+{
+	struct alias_pav_group *pos;
+	__u8 search_unit_addr;
+
+	/* for hyper pav there is only one group */
+	if (lcu->pav == HYPER_PAV) {
+		if (list_empty(&lcu->grouplist))
+			return NULL;
+		else
+			return list_first_entry(&lcu->grouplist,
+						struct alias_pav_group, group);
+	}
+
+	/* for base pav we have to find the group that matches the base */
+	if (uid->type == UA_BASE_DEVICE)
+		search_unit_addr = uid->real_unit_addr;
+	else
+		search_unit_addr = uid->base_unit_addr;
+	list_for_each_entry(pos, &lcu->grouplist, group) {
+		if (pos->uid.base_unit_addr == search_unit_addr &&
+		    !strncmp(pos->uid.vduit, uid->vduit, sizeof(uid->vduit)))
+			return pos;
+	};
+	return NULL;
+}
+
+static struct alias_server *_allocate_server(struct dasd_uid *uid)
+{
+	struct alias_server *server;
+
+	server = kzalloc(sizeof(*server), GFP_KERNEL);
+	if (!server)
+		return ERR_PTR(-ENOMEM);
+	memcpy(server->uid.vendor, uid->vendor, sizeof(uid->vendor));
+	memcpy(server->uid.serial, uid->serial, sizeof(uid->serial));
+	INIT_LIST_HEAD(&server->server);
+	INIT_LIST_HEAD(&server->lculist);
+	return server;
+}
+
+static void _free_server(struct alias_server *server)
+{
+	kfree(server);
+}
+
+static struct alias_lcu *_allocate_lcu(struct dasd_uid *uid)
+{
+	struct alias_lcu *lcu;
+
+	lcu = kzalloc(sizeof(*lcu), GFP_KERNEL);
+	if (!lcu)
+		return ERR_PTR(-ENOMEM);
+	lcu->uac = kzalloc(sizeof(*(lcu->uac)), GFP_KERNEL | GFP_DMA);
+	if (!lcu->uac)
+		goto out_err1;
+	lcu->rsu_cqr = kzalloc(sizeof(*lcu->rsu_cqr), GFP_KERNEL | GFP_DMA);
+	if (!lcu->rsu_cqr)
+		goto out_err2;
+	lcu->rsu_cqr->cpaddr = kzalloc(sizeof(struct ccw1),
+				       GFP_KERNEL | GFP_DMA);
+	if (!lcu->rsu_cqr->cpaddr)
+		goto out_err3;
+	lcu->rsu_cqr->data = kzalloc(16, GFP_KERNEL | GFP_DMA);
+	if (!lcu->rsu_cqr->data)
+		goto out_err4;
+
+	memcpy(lcu->uid.vendor, uid->vendor, sizeof(uid->vendor));
+	memcpy(lcu->uid.serial, uid->serial, sizeof(uid->serial));
+	lcu->uid.ssid = uid->ssid;
+	lcu->pav = NO_PAV;
+	lcu->flags = NEED_UAC_UPDATE | UPDATE_PENDING;
+	INIT_LIST_HEAD(&lcu->lcu);
+	INIT_LIST_HEAD(&lcu->inactive_devices);
+	INIT_LIST_HEAD(&lcu->active_devices);
+	INIT_LIST_HEAD(&lcu->grouplist);
+	INIT_WORK(&lcu->suc_data.worker, summary_unit_check_handling_work);
+	INIT_DELAYED_WORK(&lcu->ruac_data.dwork, lcu_update_work);
+	spin_lock_init(&lcu->lock);
+	init_completion(&lcu->lcu_setup);
+	return lcu;
+
+out_err4:
+	kfree(lcu->rsu_cqr->cpaddr);
+out_err3:
+	kfree(lcu->rsu_cqr);
+out_err2:
+	kfree(lcu->uac);
+out_err1:
+	kfree(lcu);
+	return ERR_PTR(-ENOMEM);
+}
+
+static void _free_lcu(struct alias_lcu *lcu)
+{
+	kfree(lcu->rsu_cqr->data);
+	kfree(lcu->rsu_cqr->cpaddr);
+	kfree(lcu->rsu_cqr);
+	kfree(lcu->uac);
+	kfree(lcu);
+}
+
+/*
+ * This is the function that will allocate all the server and lcu data,
+ * so this function must be called first for a new device.
+ * If the return value is 1, the lcu was already known before, if it
+ * is 0, this is a new lcu.
+ * Negative return code indicates that something went wrong (e.g. -ENOMEM)
+ */
+int dasd_alias_make_device_known_to_lcu(struct dasd_device *device)
+{
+	struct dasd_eckd_private *private;
+	unsigned long flags;
+	struct alias_server *server, *newserver;
+	struct alias_lcu *lcu, *newlcu;
+	struct dasd_uid uid;
+
+	private = (struct dasd_eckd_private *) device->private;
+
+	device->discipline->get_uid(device, &uid);
+	spin_lock_irqsave(&aliastree.lock, flags);
+	server = _find_server(&uid);
+	if (!server) {
+		spin_unlock_irqrestore(&aliastree.lock, flags);
+		newserver = _allocate_server(&uid);
+		if (IS_ERR(newserver))
+			return PTR_ERR(newserver);
+		spin_lock_irqsave(&aliastree.lock, flags);
+		server = _find_server(&uid);
+		if (!server) {
+			list_add(&newserver->server, &aliastree.serverlist);
+			server = newserver;
+		} else {
+			/* someone was faster */
+			_free_server(newserver);
+		}
+	}
+
+	lcu = _find_lcu(server, &uid);
+	if (!lcu) {
+		spin_unlock_irqrestore(&aliastree.lock, flags);
+		newlcu = _allocate_lcu(&uid);
+		if (IS_ERR(newlcu))
+			return PTR_ERR(newlcu);
+		spin_lock_irqsave(&aliastree.lock, flags);
+		lcu = _find_lcu(server, &uid);
+		if (!lcu) {
+			list_add(&newlcu->lcu, &server->lculist);
+			lcu = newlcu;
+		} else {
+			/* someone was faster */
+			_free_lcu(newlcu);
+		}
+	}
+	spin_lock(&lcu->lock);
+	list_add(&device->alias_list, &lcu->inactive_devices);
+	private->lcu = lcu;
+	spin_unlock(&lcu->lock);
+	spin_unlock_irqrestore(&aliastree.lock, flags);
+
+	return 0;
+}
+
+/*
+ * This function removes a device from the scope of alias management.
+ * The complicated part is to make sure that it is not in use by
+ * any of the workers. If necessary cancel the work.
+ */
+void dasd_alias_disconnect_device_from_lcu(struct dasd_device *device)
+{
+	struct dasd_eckd_private *private;
+	unsigned long flags;
+	struct alias_lcu *lcu;
+	struct alias_server *server;
+	int was_pending;
+	struct dasd_uid uid;
+
+	private = (struct dasd_eckd_private *) device->private;
+	lcu = private->lcu;
+	/* nothing to do if already disconnected */
+	if (!lcu)
+		return;
+	device->discipline->get_uid(device, &uid);
+	spin_lock_irqsave(&lcu->lock, flags);
+	list_del_init(&device->alias_list);
+	/* make sure that the workers don't use this device */
+	if (device == lcu->suc_data.device) {
+		spin_unlock_irqrestore(&lcu->lock, flags);
+		cancel_work_sync(&lcu->suc_data.worker);
+		spin_lock_irqsave(&lcu->lock, flags);
+		if (device == lcu->suc_data.device)
+			lcu->suc_data.device = NULL;
+	}
+	was_pending = 0;
+	if (device == lcu->ruac_data.device) {
+		spin_unlock_irqrestore(&lcu->lock, flags);
+		was_pending = 1;
+		cancel_delayed_work_sync(&lcu->ruac_data.dwork);
+		spin_lock_irqsave(&lcu->lock, flags);
+		if (device == lcu->ruac_data.device)
+			lcu->ruac_data.device = NULL;
+	}
+	private->lcu = NULL;
+	spin_unlock_irqrestore(&lcu->lock, flags);
+
+	spin_lock_irqsave(&aliastree.lock, flags);
+	spin_lock(&lcu->lock);
+	if (list_empty(&lcu->grouplist) &&
+	    list_empty(&lcu->active_devices) &&
+	    list_empty(&lcu->inactive_devices)) {
+		list_del(&lcu->lcu);
+		spin_unlock(&lcu->lock);
+		_free_lcu(lcu);
+		lcu = NULL;
+	} else {
+		if (was_pending)
+			_schedule_lcu_update(lcu, NULL);
+		spin_unlock(&lcu->lock);
+	}
+	server = _find_server(&uid);
+	if (server && list_empty(&server->lculist)) {
+		list_del(&server->server);
+		_free_server(server);
+	}
+	spin_unlock_irqrestore(&aliastree.lock, flags);
+}
+
+/*
+ * This function assumes that the unit address configuration stored
+ * in the lcu is up to date and will update the device uid before
+ * adding it to a pav group.
+ */
+
+static int _add_device_to_lcu(struct alias_lcu *lcu,
+			      struct dasd_device *device,
+			      struct dasd_device *pos)
+{
+
+	struct dasd_eckd_private *private;
+	struct alias_pav_group *group;
+	struct dasd_uid uid;
+	unsigned long flags;
+
+	private = (struct dasd_eckd_private *) device->private;
+
+	/* only lock if not already locked */
+	if (device != pos)
+		spin_lock_irqsave_nested(get_ccwdev_lock(device->cdev), flags,
+					 CDEV_NESTED_SECOND);
+	private->uid.type = lcu->uac->unit[private->uid.real_unit_addr].ua_type;
+	private->uid.base_unit_addr =
+		lcu->uac->unit[private->uid.real_unit_addr].base_ua;
+	uid = private->uid;
+
+	if (device != pos)
+		spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
+
+	/* if we have no PAV anyway, we don't need to bother with PAV groups */
+	if (lcu->pav == NO_PAV) {
+		list_move(&device->alias_list, &lcu->active_devices);
+		return 0;
+	}
+
+	group = _find_group(lcu, &uid);
+	if (!group) {
+		group = kzalloc(sizeof(*group), GFP_ATOMIC);
+		if (!group)
+			return -ENOMEM;
+		memcpy(group->uid.vendor, uid.vendor, sizeof(uid.vendor));
+		memcpy(group->uid.serial, uid.serial, sizeof(uid.serial));
+		group->uid.ssid = uid.ssid;
+		if (uid.type == UA_BASE_DEVICE)
+			group->uid.base_unit_addr = uid.real_unit_addr;
+		else
+			group->uid.base_unit_addr = uid.base_unit_addr;
+		memcpy(group->uid.vduit, uid.vduit, sizeof(uid.vduit));
+		INIT_LIST_HEAD(&group->group);
+		INIT_LIST_HEAD(&group->baselist);
+		INIT_LIST_HEAD(&group->aliaslist);
+		list_add(&group->group, &lcu->grouplist);
+	}
+	if (uid.type == UA_BASE_DEVICE)
+		list_move(&device->alias_list, &group->baselist);
+	else
+		list_move(&device->alias_list, &group->aliaslist);
+	private->pavgroup = group;
+	return 0;
+};
+
+static void _remove_device_from_lcu(struct alias_lcu *lcu,
+				    struct dasd_device *device)
+{
+	struct dasd_eckd_private *private;
+	struct alias_pav_group *group;
+
+	private = (struct dasd_eckd_private *) device->private;
+	list_move(&device->alias_list, &lcu->inactive_devices);
+	group = private->pavgroup;
+	if (!group)
+		return;
+	private->pavgroup = NULL;
+	if (list_empty(&group->baselist) && list_empty(&group->aliaslist)) {
+		list_del(&group->group);
+		kfree(group);
+		return;
+	}
+	if (group->next == device)
+		group->next = NULL;
+};
+
+static int
+suborder_not_supported(struct dasd_ccw_req *cqr)
+{
+	char *sense;
+	char reason;
+	char msg_format;
+	char msg_no;
+
+	sense = dasd_get_sense(&cqr->irb);
+	if (!sense)
+		return 0;
+
+	reason = sense[0];
+	msg_format = (sense[7] & 0xF0);
+	msg_no = (sense[7] & 0x0F);
+
+	/* command reject, Format 0 MSG 4 - invalid parameter */
+	if ((reason == 0x80) && (msg_format == 0x00) && (msg_no == 0x04))
+		return 1;
+
+	return 0;
+}
+
+static int read_unit_address_configuration(struct dasd_device *device,
+					   struct alias_lcu *lcu)
+{
+	struct dasd_psf_prssd_data *prssdp;
+	struct dasd_ccw_req *cqr;
+	struct ccw1 *ccw;
+	int rc;
+	unsigned long flags;
+
+	cqr = dasd_kmalloc_request(DASD_ECKD_MAGIC, 1 /* PSF */	+ 1 /* RSSD */,
+				   (sizeof(struct dasd_psf_prssd_data)),
+				   device);
+	if (IS_ERR(cqr))
+		return PTR_ERR(cqr);
+	cqr->startdev = device;
+	cqr->memdev = device;
+	clear_bit(DASD_CQR_FLAGS_USE_ERP, &cqr->flags);
+	cqr->retries = 10;
+	cqr->expires = 20 * HZ;
+
+	/* Prepare for Read Subsystem Data */
+	prssdp = (struct dasd_psf_prssd_data *) cqr->data;
+	memset(prssdp, 0, sizeof(struct dasd_psf_prssd_data));
+	prssdp->order = PSF_ORDER_PRSSD;
+	prssdp->suborder = 0x0e;	/* Read unit address configuration */
+	/* all other bytes of prssdp must be zero */
+
+	ccw = cqr->cpaddr;
+	ccw->cmd_code = DASD_ECKD_CCW_PSF;
+	ccw->count = sizeof(struct dasd_psf_prssd_data);
+	ccw->flags |= CCW_FLAG_CC;
+	ccw->cda = (__u32)(addr_t) prssdp;
+
+	/* Read Subsystem Data - feature codes */
+	memset(lcu->uac, 0, sizeof(*(lcu->uac)));
+
+	ccw++;
+	ccw->cmd_code = DASD_ECKD_CCW_RSSD;
+	ccw->count = sizeof(*(lcu->uac));
+	ccw->cda = (__u32)(addr_t) lcu->uac;
+
+	cqr->buildclk = get_tod_clock();
+	cqr->status = DASD_CQR_FILLED;
+
+	/* need to unset flag here to detect race with summary unit check */
+	spin_lock_irqsave(&lcu->lock, flags);
+	lcu->flags &= ~NEED_UAC_UPDATE;
+	spin_unlock_irqrestore(&lcu->lock, flags);
+
+	do {
+		rc = dasd_sleep_on(cqr);
+		if (rc && suborder_not_supported(cqr))
+			return -EOPNOTSUPP;
+	} while (rc && (cqr->retries > 0));
+	if (rc) {
+		spin_lock_irqsave(&lcu->lock, flags);
+		lcu->flags |= NEED_UAC_UPDATE;
+		spin_unlock_irqrestore(&lcu->lock, flags);
+	}
+	dasd_kfree_request(cqr, cqr->memdev);
+	return rc;
+}
+
+static int _lcu_update(struct dasd_device *refdev, struct alias_lcu *lcu)
+{
+	unsigned long flags;
+	struct alias_pav_group *pavgroup, *tempgroup;
+	struct dasd_device *device, *tempdev;
+	int i, rc;
+	struct dasd_eckd_private *private;
+
+	spin_lock_irqsave(&lcu->lock, flags);
+	list_for_each_entry_safe(pavgroup, tempgroup, &lcu->grouplist, group) {
+		list_for_each_entry_safe(device, tempdev, &pavgroup->baselist,
+					 alias_list) {
+			list_move(&device->alias_list, &lcu->active_devices);
+			private = (struct dasd_eckd_private *) device->private;
+			private->pavgroup = NULL;
+		}
+		list_for_each_entry_safe(device, tempdev, &pavgroup->aliaslist,
+					 alias_list) {
+			list_move(&device->alias_list, &lcu->active_devices);
+			private = (struct dasd_eckd_private *) device->private;
+			private->pavgroup = NULL;
+		}
+		list_del(&pavgroup->group);
+		kfree(pavgroup);
+	}
+	spin_unlock_irqrestore(&lcu->lock, flags);
+
+	rc = read_unit_address_configuration(refdev, lcu);
+	if (rc)
+		return rc;
+
+	/* need to take cdev lock before lcu lock */
+	spin_lock_irqsave_nested(get_ccwdev_lock(refdev->cdev), flags,
+				 CDEV_NESTED_FIRST);
+	spin_lock(&lcu->lock);
+	lcu->pav = NO_PAV;
+	for (i = 0; i < MAX_DEVICES_PER_LCU; ++i) {
+		switch (lcu->uac->unit[i].ua_type) {
+		case UA_BASE_PAV_ALIAS:
+			lcu->pav = BASE_PAV;
+			break;
+		case UA_HYPER_PAV_ALIAS:
+			lcu->pav = HYPER_PAV;
+			break;
+		}
+		if (lcu->pav != NO_PAV)
+			break;
+	}
+
+	list_for_each_entry_safe(device, tempdev, &lcu->active_devices,
+				 alias_list) {
+		_add_device_to_lcu(lcu, device, refdev);
+	}
+	spin_unlock(&lcu->lock);
+	spin_unlock_irqrestore(get_ccwdev_lock(refdev->cdev), flags);
+	return 0;
+}
+
+static void lcu_update_work(struct work_struct *work)
+{
+	struct alias_lcu *lcu;
+	struct read_uac_work_data *ruac_data;
+	struct dasd_device *device;
+	unsigned long flags;
+	int rc;
+
+	ruac_data = container_of(work, struct read_uac_work_data, dwork.work);
+	lcu = container_of(ruac_data, struct alias_lcu, ruac_data);
+	device = ruac_data->device;
+	rc = _lcu_update(device, lcu);
+	/*
+	 * Need to check flags again, as there could have been another
+	 * prepare_update or a new device a new device while we were still
+	 * processing the data
+	 */
+	spin_lock_irqsave(&lcu->lock, flags);
+	if ((rc && (rc != -EOPNOTSUPP)) || (lcu->flags & NEED_UAC_UPDATE)) {
+		DBF_DEV_EVENT(DBF_WARNING, device, "could not update"
+			    " alias data in lcu (rc = %d), retry later", rc);
+		schedule_delayed_work(&lcu->ruac_data.dwork, 30*HZ);
+	} else {
+		lcu->ruac_data.device = NULL;
+		lcu->flags &= ~UPDATE_PENDING;
+	}
+	spin_unlock_irqrestore(&lcu->lock, flags);
+}
+
+static int _schedule_lcu_update(struct alias_lcu *lcu,
+				struct dasd_device *device)
+{
+	struct dasd_device *usedev = NULL;
+	struct alias_pav_group *group;
+
+	lcu->flags |= NEED_UAC_UPDATE;
+	if (lcu->ruac_data.device) {
+		/* already scheduled or running */
+		return 0;
+	}
+	if (device && !list_empty(&device->alias_list))
+		usedev = device;
+
+	if (!usedev && !list_empty(&lcu->grouplist)) {
+		group = list_first_entry(&lcu->grouplist,
+					 struct alias_pav_group, group);
+		if (!list_empty(&group->baselist))
+			usedev = list_first_entry(&group->baselist,
+						  struct dasd_device,
+						  alias_list);
+		else if (!list_empty(&group->aliaslist))
+			usedev = list_first_entry(&group->aliaslist,
+						  struct dasd_device,
+						  alias_list);
+	}
+	if (!usedev && !list_empty(&lcu->active_devices)) {
+		usedev = list_first_entry(&lcu->active_devices,
+					  struct dasd_device, alias_list);
+	}
+	/*
+	 * if we haven't found a proper device yet, give up for now, the next
+	 * device that will be set active will trigger an lcu update
+	 */
+	if (!usedev)
+		return -EINVAL;
+	lcu->ruac_data.device = usedev;
+	schedule_delayed_work(&lcu->ruac_data.dwork, 0);
+	return 0;
+}
+
+int dasd_alias_add_device(struct dasd_device *device)
+{
+	struct dasd_eckd_private *private;
+	struct alias_lcu *lcu;
+	unsigned long flags;
+	int rc;
+
+	private = (struct dasd_eckd_private *) device->private;
+	lcu = private->lcu;
+	rc = 0;
+
+	/* need to take cdev lock before lcu lock */
+	spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+	spin_lock(&lcu->lock);
+	if (!(lcu->flags & UPDATE_PENDING)) {
+		rc = _add_device_to_lcu(lcu, device, device);
+		if (rc)
+			lcu->flags |= UPDATE_PENDING;
+	}
+	if (lcu->flags & UPDATE_PENDING) {
+		list_move(&device->alias_list, &lcu->active_devices);
+		_schedule_lcu_update(lcu, device);
+	}
+	spin_unlock(&lcu->lock);
+	spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
+	return rc;
+}
+
+int dasd_alias_update_add_device(struct dasd_device *device)
+{
+	struct dasd_eckd_private *private;
+	private = (struct dasd_eckd_private *) device->private;
+	private->lcu->flags |= UPDATE_PENDING;
+	return dasd_alias_add_device(device);
+}
+
+int dasd_alias_remove_device(struct dasd_device *device)
+{
+	struct dasd_eckd_private *private;
+	struct alias_lcu *lcu;
+	unsigned long flags;
+
+	private = (struct dasd_eckd_private *) device->private;
+	lcu = private->lcu;
+	/* nothing to do if already removed */
+	if (!lcu)
+		return 0;
+	spin_lock_irqsave(&lcu->lock, flags);
+	_remove_device_from_lcu(lcu, device);
+	spin_unlock_irqrestore(&lcu->lock, flags);
+	return 0;
+}
+
+struct dasd_device *dasd_alias_get_start_dev(struct dasd_device *base_device)
+{
+
+	struct dasd_device *alias_device;
+	struct alias_pav_group *group;
+	struct alias_lcu *lcu;
+	struct dasd_eckd_private *private, *alias_priv;
+	unsigned long flags;
+
+	private = (struct dasd_eckd_private *) base_device->private;
+	group = private->pavgroup;
+	lcu = private->lcu;
+	if (!group || !lcu)
+		return NULL;
+	if (lcu->pav == NO_PAV ||
+	    lcu->flags & (NEED_UAC_UPDATE | UPDATE_PENDING))
+		return NULL;
+	if (unlikely(!(private->features.feature[8] & 0x01))) {
+		/*
+		 * PAV enabled but prefix not, very unlikely
+		 * seems to be a lost pathgroup
+		 * use base device to do IO
+		 */
+		DBF_DEV_EVENT(DBF_ERR, base_device, "%s",
+			      "Prefix not enabled with PAV enabled\n");
+		return NULL;
+	}
+
+	spin_lock_irqsave(&lcu->lock, flags);
+	alias_device = group->next;
+	if (!alias_device) {
+		if (list_empty(&group->aliaslist)) {
+			spin_unlock_irqrestore(&lcu->lock, flags);
+			return NULL;
+		} else {
+			alias_device = list_first_entry(&group->aliaslist,
+							struct dasd_device,
+							alias_list);
+		}
+	}
+	if (list_is_last(&alias_device->alias_list, &group->aliaslist))
+		group->next = list_first_entry(&group->aliaslist,
+					       struct dasd_device, alias_list);
+	else
+		group->next = list_first_entry(&alias_device->alias_list,
+					       struct dasd_device, alias_list);
+	spin_unlock_irqrestore(&lcu->lock, flags);
+	alias_priv = (struct dasd_eckd_private *) alias_device->private;
+	if ((alias_priv->count < private->count) && !alias_device->stopped)
+		return alias_device;
+	else
+		return NULL;
+}
+
+/*
+ * Summary unit check handling depends on the way alias devices
+ * are handled so it is done here rather then in dasd_eckd.c
+ */
+static int reset_summary_unit_check(struct alias_lcu *lcu,
+				    struct dasd_device *device,
+				    char reason)
+{
+	struct dasd_ccw_req *cqr;
+	int rc = 0;
+	struct ccw1 *ccw;
+
+	cqr = lcu->rsu_cqr;
+	strncpy((char *) &cqr->magic, "ECKD", 4);
+	ASCEBC((char *) &cqr->magic, 4);
+	ccw = cqr->cpaddr;
+	ccw->cmd_code = DASD_ECKD_CCW_RSCK;
+	ccw->flags = 0 ;
+	ccw->count = 16;
+	ccw->cda = (__u32)(addr_t) cqr->data;
+	((char *)cqr->data)[0] = reason;
+
+	clear_bit(DASD_CQR_FLAGS_USE_ERP, &cqr->flags);
+	cqr->retries = 255;	/* set retry counter to enable basic ERP */
+	cqr->startdev = device;
+	cqr->memdev = device;
+	cqr->block = NULL;
+	cqr->expires = 5 * HZ;
+	cqr->buildclk = get_tod_clock();
+	cqr->status = DASD_CQR_FILLED;
+
+	rc = dasd_sleep_on_immediatly(cqr);
+	return rc;
+}
+
+static void _restart_all_base_devices_on_lcu(struct alias_lcu *lcu)
+{
+	struct alias_pav_group *pavgroup;
+	struct dasd_device *device;
+	struct dasd_eckd_private *private;
+	unsigned long flags;
+
+	/* active and inactive list can contain alias as well as base devices */
+	list_for_each_entry(device, &lcu->active_devices, alias_list) {
+		private = (struct dasd_eckd_private *) device->private;
+		spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+		if (private->uid.type != UA_BASE_DEVICE) {
+			spin_unlock_irqrestore(get_ccwdev_lock(device->cdev),
+					       flags);
+			continue;
+		}
+		spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
+		dasd_schedule_block_bh(device->block);
+		dasd_schedule_device_bh(device);
+	}
+	list_for_each_entry(device, &lcu->inactive_devices, alias_list) {
+		private = (struct dasd_eckd_private *) device->private;
+		spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+		if (private->uid.type != UA_BASE_DEVICE) {
+			spin_unlock_irqrestore(get_ccwdev_lock(device->cdev),
+					       flags);
+			continue;
+		}
+		spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
+		dasd_schedule_block_bh(device->block);
+		dasd_schedule_device_bh(device);
+	}
+	list_for_each_entry(pavgroup, &lcu->grouplist, group) {
+		list_for_each_entry(device, &pavgroup->baselist, alias_list) {
+			dasd_schedule_block_bh(device->block);
+			dasd_schedule_device_bh(device);
+		}
+	}
+}
+
+static void flush_all_alias_devices_on_lcu(struct alias_lcu *lcu)
+{
+	struct alias_pav_group *pavgroup;
+	struct dasd_device *device, *temp;
+	struct dasd_eckd_private *private;
+	int rc;
+	unsigned long flags;
+	LIST_HEAD(active);
+
+	/*
+	 * Problem here ist that dasd_flush_device_queue may wait
+	 * for termination of a request to complete. We can't keep
+	 * the lcu lock during that time, so we must assume that
+	 * the lists may have changed.
+	 * Idea: first gather all active alias devices in a separate list,
+	 * then flush the first element of this list unlocked, and afterwards
+	 * check if it is still on the list before moving it to the
+	 * active_devices list.
+	 */
+
+	spin_lock_irqsave(&lcu->lock, flags);
+	list_for_each_entry_safe(device, temp, &lcu->active_devices,
+				 alias_list) {
+		private = (struct dasd_eckd_private *) device->private;
+		if (private->uid.type == UA_BASE_DEVICE)
+			continue;
+		list_move(&device->alias_list, &active);
+	}
+
+	list_for_each_entry(pavgroup, &lcu->grouplist, group) {
+		list_splice_init(&pavgroup->aliaslist, &active);
+	}
+	while (!list_empty(&active)) {
+		device = list_first_entry(&active, struct dasd_device,
+					  alias_list);
+		spin_unlock_irqrestore(&lcu->lock, flags);
+		rc = dasd_flush_device_queue(device);
+		spin_lock_irqsave(&lcu->lock, flags);
+		/*
+		 * only move device around if it wasn't moved away while we
+		 * were waiting for the flush
+		 */
+		if (device == list_first_entry(&active,
+					       struct dasd_device, alias_list))
+			list_move(&device->alias_list, &lcu->active_devices);
+	}
+	spin_unlock_irqrestore(&lcu->lock, flags);
+}
+
+static void __stop_device_on_lcu(struct dasd_device *device,
+				 struct dasd_device *pos)
+{
+	/* If pos == device then device is already locked! */
+	if (pos == device) {
+		dasd_device_set_stop_bits(pos, DASD_STOPPED_SU);
+		return;
+	}
+	spin_lock(get_ccwdev_lock(pos->cdev));
+	dasd_device_set_stop_bits(pos, DASD_STOPPED_SU);
+	spin_unlock(get_ccwdev_lock(pos->cdev));
+}
+
+/*
+ * This function is called in interrupt context, so the
+ * cdev lock for device is already locked!
+ */
+static void _stop_all_devices_on_lcu(struct alias_lcu *lcu,
+				     struct dasd_device *device)
+{
+	struct alias_pav_group *pavgroup;
+	struct dasd_device *pos;
+
+	list_for_each_entry(pos, &lcu->active_devices, alias_list)
+		__stop_device_on_lcu(device, pos);
+	list_for_each_entry(pos, &lcu->inactive_devices, alias_list)
+		__stop_device_on_lcu(device, pos);
+	list_for_each_entry(pavgroup, &lcu->grouplist, group) {
+		list_for_each_entry(pos, &pavgroup->baselist, alias_list)
+			__stop_device_on_lcu(device, pos);
+		list_for_each_entry(pos, &pavgroup->aliaslist, alias_list)
+			__stop_device_on_lcu(device, pos);
+	}
+}
+
+static void _unstop_all_devices_on_lcu(struct alias_lcu *lcu)
+{
+	struct alias_pav_group *pavgroup;
+	struct dasd_device *device;
+	unsigned long flags;
+
+	list_for_each_entry(device, &lcu->active_devices, alias_list) {
+		spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+		dasd_device_remove_stop_bits(device, DASD_STOPPED_SU);
+		spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
+	}
+
+	list_for_each_entry(device, &lcu->inactive_devices, alias_list) {
+		spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+		dasd_device_remove_stop_bits(device, DASD_STOPPED_SU);
+		spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
+	}
+
+	list_for_each_entry(pavgroup, &lcu->grouplist, group) {
+		list_for_each_entry(device, &pavgroup->baselist, alias_list) {
+			spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+			dasd_device_remove_stop_bits(device, DASD_STOPPED_SU);
+			spin_unlock_irqrestore(get_ccwdev_lock(device->cdev),
+					       flags);
+		}
+		list_for_each_entry(device, &pavgroup->aliaslist, alias_list) {
+			spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+			dasd_device_remove_stop_bits(device, DASD_STOPPED_SU);
+			spin_unlock_irqrestore(get_ccwdev_lock(device->cdev),
+					       flags);
+		}
+	}
+}
+
+static void summary_unit_check_handling_work(struct work_struct *work)
+{
+	struct alias_lcu *lcu;
+	struct summary_unit_check_work_data *suc_data;
+	unsigned long flags;
+	struct dasd_device *device;
+
+	suc_data = container_of(work, struct summary_unit_check_work_data,
+				worker);
+	lcu = container_of(suc_data, struct alias_lcu, suc_data);
+	device = suc_data->device;
+
+	/* 1. flush alias devices */
+	flush_all_alias_devices_on_lcu(lcu);
+
+	/* 2. reset summary unit check */
+	spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+	dasd_device_remove_stop_bits(device,
+				     (DASD_STOPPED_SU | DASD_STOPPED_PENDING));
+	spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
+	reset_summary_unit_check(lcu, device, suc_data->reason);
+
+	spin_lock_irqsave(&lcu->lock, flags);
+	_unstop_all_devices_on_lcu(lcu);
+	_restart_all_base_devices_on_lcu(lcu);
+	/* 3. read new alias configuration */
+	_schedule_lcu_update(lcu, device);
+	lcu->suc_data.device = NULL;
+	spin_unlock_irqrestore(&lcu->lock, flags);
+}
+
+/*
+ * note: this will be called from int handler context (cdev locked)
+ */
+void dasd_alias_handle_summary_unit_check(struct dasd_device *device,
+					  struct irb *irb)
+{
+	struct alias_lcu *lcu;
+	char reason;
+	struct dasd_eckd_private *private;
+	char *sense;
+
+	private = (struct dasd_eckd_private *) device->private;
+
+	sense = dasd_get_sense(irb);
+	if (sense) {
+		reason = sense[8];
+		DBF_DEV_EVENT(DBF_NOTICE, device, "%s %x",
+			    "eckd handle summary unit check: reason", reason);
+	} else {
+		DBF_DEV_EVENT(DBF_WARNING, device, "%s",
+			    "eckd handle summary unit check:"
+			    " no reason code available");
+		return;
+	}
+
+	lcu = private->lcu;
+	if (!lcu) {
+		DBF_DEV_EVENT(DBF_WARNING, device, "%s",
+			    "device not ready to handle summary"
+			    " unit check (no lcu structure)");
+		return;
+	}
+	spin_lock(&lcu->lock);
+	_stop_all_devices_on_lcu(lcu, device);
+	/* prepare for lcu_update */
+	private->lcu->flags |= NEED_UAC_UPDATE | UPDATE_PENDING;
+	/* If this device is about to be removed just return and wait for
+	 * the next interrupt on a different device
+	 */
+	if (list_empty(&device->alias_list)) {
+		DBF_DEV_EVENT(DBF_WARNING, device, "%s",
+			    "device is in offline processing,"
+			    " don't do summary unit check handling");
+		spin_unlock(&lcu->lock);
+		return;
+	}
+	if (lcu->suc_data.device) {
+		/* already scheduled or running */
+		DBF_DEV_EVENT(DBF_WARNING, device, "%s",
+			    "previous instance of summary unit check worker"
+			    " still pending");
+		spin_unlock(&lcu->lock);
+		return ;
+	}
+	lcu->suc_data.reason = reason;
+	lcu->suc_data.device = device;
+	spin_unlock(&lcu->lock);
+	schedule_work(&lcu->suc_data.worker);
+};

drivers/s390/block/dasd_diag.c

@@ -0,0 +1,663 @@
+/*
+ * Author(s)......: Holger Smolinski <Holger.Smolinski@de.ibm.com>
+ * Based on.......: linux/drivers/s390/block/mdisk.c
+ * ...............: by Hartmunt Penner <hpenner@de.ibm.com>
+ * Bugreports.to..: <Linux390@de.ibm.com>
+ * Copyright IBM Corp. 1999, 2000
+ *
+ */
+
+#define KMSG_COMPONENT "dasd"
+
+#include <linux/kernel_stat.h>
+#include <linux/stddef.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/hdreg.h>
+#include <linux/bio.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/jiffies.h>
+
+#include <asm/dasd.h>
+#include <asm/debug.h>
+#include <asm/ebcdic.h>
+#include <asm/io.h>
+#include <asm/irq.h>
+#include <asm/vtoc.h>
+#include <asm/diag.h>
+
+#include "dasd_int.h"
+#include "dasd_diag.h"
+
+#define PRINTK_HEADER "dasd(diag):"
+
+MODULE_LICENSE("GPL");
+
+/* The maximum number of blocks per request (max_blocks) is dependent on the
+ * amount of storage that is available in the static I/O buffer for each
+ * device. Currently each device gets 2 pages. We want to fit two requests
+ * into the available memory so that we can immediately start the next if one
+ * finishes. */
+#define DIAG_MAX_BLOCKS	(((2 * PAGE_SIZE - sizeof(struct dasd_ccw_req) - \
+			   sizeof(struct dasd_diag_req)) / \
+		           sizeof(struct dasd_diag_bio)) / 2)
+#define DIAG_MAX_RETRIES	32
+#define DIAG_TIMEOUT		50
+
+static struct dasd_discipline dasd_diag_discipline;
+
+struct dasd_diag_private {
+	struct dasd_diag_characteristics rdc_data;
+	struct dasd_diag_rw_io iob;
+	struct dasd_diag_init_io iib;
+	blocknum_t pt_block;
+	struct ccw_dev_id dev_id;
+};
+
+struct dasd_diag_req {
+	unsigned int block_count;
+	struct dasd_diag_bio bio[0];
+};
+
+static const u8 DASD_DIAG_CMS1[] = { 0xc3, 0xd4, 0xe2, 0xf1 };/* EBCDIC CMS1 */
+
+/* Perform DIAG250 call with block I/O parameter list iob (input and output)
+ * and function code cmd.
+ * In case of an exception return 3. Otherwise return result of bitwise OR of
+ * resulting condition code and DIAG return code. */
+static inline int dia250(void *iob, int cmd)
+{
+	register unsigned long reg2 asm ("2") = (unsigned long) iob;
+	typedef union {
+		struct dasd_diag_init_io init_io;
+		struct dasd_diag_rw_io rw_io;
+	} addr_type;
+	int rc;
+
+	rc = 3;
+	asm volatile(
+		"	diag	2,%2,0x250\n"
+		"0:	ipm	%0\n"
+		"	srl	%0,28\n"
+		"	or	%0,3\n"
+		"1:\n"
+		EX_TABLE(0b,1b)
+		: "+d" (rc), "=m" (*(addr_type *) iob)
+		: "d" (cmd), "d" (reg2), "m" (*(addr_type *) iob)
+		: "3", "cc");
+	return rc;
+}
+
+/* Initialize block I/O to DIAG device using the specified blocksize and
+ * block offset. On success, return zero and set end_block to contain the
+ * number of blocks on the device minus the specified offset. Return non-zero
+ * otherwise. */
+static inline int
+mdsk_init_io(struct dasd_device *device, unsigned int blocksize,
+	     blocknum_t offset, blocknum_t *end_block)
+{
+	struct dasd_diag_private *private;
+	struct dasd_diag_init_io *iib;
+	int rc;
+
+	private = (struct dasd_diag_private *) device->private;
+	iib = &private->iib;
+	memset(iib, 0, sizeof (struct dasd_diag_init_io));
+
+	iib->dev_nr = private->dev_id.devno;
+	iib->block_size = blocksize;
+	iib->offset = offset;
+	iib->flaga = DASD_DIAG_FLAGA_DEFAULT;
+
+	rc = dia250(iib, INIT_BIO);
+
+	if ((rc & 3) == 0 && end_block)
+		*end_block = iib->end_block;
+
+	return rc;
+}
+
+/* Remove block I/O environment for device. Return zero on success, non-zero
+ * otherwise. */
+static inline int
+mdsk_term_io(struct dasd_device * device)
+{
+	struct dasd_diag_private *private;
+	struct dasd_diag_init_io *iib;
+	int rc;
+
+	private = (struct dasd_diag_private *) device->private;
+	iib = &private->iib;
+	memset(iib, 0, sizeof (struct dasd_diag_init_io));
+	iib->dev_nr = private->dev_id.devno;
+	rc = dia250(iib, TERM_BIO);
+	return rc;
+}
+
+/* Error recovery for failed DIAG requests - try to reestablish the DIAG
+ * environment. */
+static void
+dasd_diag_erp(struct dasd_device *device)
+{
+	int rc;
+
+	mdsk_term_io(device);
+	rc = mdsk_init_io(device, device->block->bp_block, 0, NULL);
+	if (rc == 4) {
+		if (!(test_and_set_bit(DASD_FLAG_DEVICE_RO, &device->flags)))
+			pr_warning("%s: The access mode of a DIAG device "
+				   "changed to read-only\n",
+				   dev_name(&device->cdev->dev));
+		rc = 0;
+	}
+	if (rc)
+		pr_warning("%s: DIAG ERP failed with "
+			    "rc=%d\n", dev_name(&device->cdev->dev), rc);
+}
+
+/* Start a given request at the device. Return zero on success, non-zero
+ * otherwise. */
+static int
+dasd_start_diag(struct dasd_ccw_req * cqr)
+{
+	struct dasd_device *device;
+	struct dasd_diag_private *private;
+	struct dasd_diag_req *dreq;
+	int rc;
+
+	device = cqr->startdev;
+	if (cqr->retries < 0) {
+		DBF_DEV_EVENT(DBF_ERR, device, "DIAG start_IO: request %p "
+			    "- no retry left)", cqr);
+		cqr->status = DASD_CQR_ERROR;
+		return -EIO;
+	}
+	private = (struct dasd_diag_private *) device->private;
+	dreq = (struct dasd_diag_req *) cqr->data;
+
+	private->iob.dev_nr = private->dev_id.devno;
+	private->iob.key = 0;
+	private->iob.flags = DASD_DIAG_RWFLAG_ASYNC;
+	private->iob.block_count = dreq->block_count;
+	private->iob.interrupt_params = (addr_t) cqr;
+	private->iob.bio_list = dreq->bio;
+	private->iob.flaga = DASD_DIAG_FLAGA_DEFAULT;
+
+	cqr->startclk = get_tod_clock();
+	cqr->starttime = jiffies;
+	cqr->retries--;
+
+	rc = dia250(&private->iob, RW_BIO);
+	switch (rc) {
+	case 0: /* Synchronous I/O finished successfully */
+		cqr->stopclk = get_tod_clock();
+		cqr->status = DASD_CQR_SUCCESS;
+		/* Indicate to calling function that only a dasd_schedule_bh()
+		   and no timer is needed */
+                rc = -EACCES;
+		break;
+	case 8: /* Asynchronous I/O was started */
+		cqr->status = DASD_CQR_IN_IO;
+		rc = 0;
+		break;
+	default: /* Error condition */
+		cqr->status = DASD_CQR_QUEUED;
+		DBF_DEV_EVENT(DBF_WARNING, device, "dia250 returned rc=%d", rc);
+		dasd_diag_erp(device);
+		rc = -EIO;
+		break;
+	}
+	cqr->intrc = rc;
+	return rc;
+}
+
+/* Terminate given request at the device. */
+static int
+dasd_diag_term_IO(struct dasd_ccw_req * cqr)
+{
+	struct dasd_device *device;
+
+	device = cqr->startdev;
+	mdsk_term_io(device);
+	mdsk_init_io(device, device->block->bp_block, 0, NULL);
+	cqr->status = DASD_CQR_CLEAR_PENDING;
+	cqr->stopclk = get_tod_clock();
+	dasd_schedule_device_bh(device);
+	return 0;
+}
+
+/* Handle external interruption. */
+static void dasd_ext_handler(struct ext_code ext_code,
+			     unsigned int param32, unsigned long param64)
+{
+	struct dasd_ccw_req *cqr, *next;
+	struct dasd_device *device;
+	unsigned long long expires;
+	unsigned long flags;
+	addr_t ip;
+	int rc;
+
+	switch (ext_code.subcode >> 8) {
+	case DASD_DIAG_CODE_31BIT:
+		ip = (addr_t) param32;
+		break;
+	case DASD_DIAG_CODE_64BIT:
+		ip = (addr_t) param64;
+		break;
+	default:
+		return;
+	}
+	inc_irq_stat(IRQEXT_DSD);
+	if (!ip) {		/* no intparm: unsolicited interrupt */
+		DBF_EVENT(DBF_NOTICE, "%s", "caught unsolicited "
+			      "interrupt");
+		return;
+	}
+	cqr = (struct dasd_ccw_req *) ip;
+	device = (struct dasd_device *) cqr->startdev;
+	if (strncmp(device->discipline->ebcname, (char *) &cqr->magic, 4)) {
+		DBF_DEV_EVENT(DBF_WARNING, device,
+			    " magic number of dasd_ccw_req 0x%08X doesn't"
+			    " match discipline 0x%08X",
+			    cqr->magic, *(int *) (&device->discipline->name));
+		return;
+	}
+
+	/* get irq lock to modify request queue */
+	spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+
+	/* Check for a pending clear operation */
+	if (cqr->status == DASD_CQR_CLEAR_PENDING) {
+		cqr->status = DASD_CQR_CLEARED;
+		dasd_device_clear_timer(device);
+		dasd_schedule_device_bh(device);
+		spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
+		return;
+	}
+
+	cqr->stopclk = get_tod_clock();
+
+	expires = 0;
+	if ((ext_code.subcode & 0xff) == 0) {
+		cqr->status = DASD_CQR_SUCCESS;
+		/* Start first request on queue if possible -> fast_io. */
+		if (!list_empty(&device->ccw_queue)) {
+			next = list_entry(device->ccw_queue.next,
+					  struct dasd_ccw_req, devlist);
+			if (next->status == DASD_CQR_QUEUED) {
+				rc = dasd_start_diag(next);
+				if (rc == 0)
+					expires = next->expires;
+			}
+		}
+	} else {
+		cqr->status = DASD_CQR_QUEUED;
+		DBF_DEV_EVENT(DBF_DEBUG, device, "interrupt status for "
+			      "request %p was %d (%d retries left)", cqr,
+			      ext_code.subcode & 0xff, cqr->retries);
+		dasd_diag_erp(device);
+	}
+
+	if (expires != 0)
+		dasd_device_set_timer(device, expires);
+	else
+		dasd_device_clear_timer(device);
+	dasd_schedule_device_bh(device);
+
+	spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
+}
+
+/* Check whether device can be controlled by DIAG discipline. Return zero on
+ * success, non-zero otherwise. */
+static int
+dasd_diag_check_device(struct dasd_device *device)
+{
+	struct dasd_block *block;
+	struct dasd_diag_private *private;
+	struct dasd_diag_characteristics *rdc_data;
+	struct dasd_diag_bio bio;
+	struct vtoc_cms_label *label;
+	blocknum_t end_block;
+	unsigned int sb, bsize;
+	int rc;
+
+	private = (struct dasd_diag_private *) device->private;
+	if (private == NULL) {
+		private = kzalloc(sizeof(struct dasd_diag_private),GFP_KERNEL);
+		if (private == NULL) {
+			DBF_DEV_EVENT(DBF_WARNING, device, "%s",
+				"Allocating memory for private DASD data "
+				      "failed\n");
+			return -ENOMEM;
+		}
+		ccw_device_get_id(device->cdev, &private->dev_id);
+		device->private = (void *) private;
+	}
+	block = dasd_alloc_block();
+	if (IS_ERR(block)) {
+		DBF_DEV_EVENT(DBF_WARNING, device, "%s",
+			    "could not allocate dasd block structure");
+		device->private = NULL;
+		kfree(private);
+		return PTR_ERR(block);
+	}
+	device->block = block;
+	block->base = device;
+
+	/* Read Device Characteristics */
+	rdc_data = (void *) &(private->rdc_data);
+	rdc_data->dev_nr = private->dev_id.devno;
+	rdc_data->rdc_len = sizeof (struct dasd_diag_characteristics);
+
+	rc = diag210((struct diag210 *) rdc_data);
+	if (rc) {
+		DBF_DEV_EVENT(DBF_WARNING, device, "failed to retrieve device "
+			    "information (rc=%d)", rc);
+		rc = -EOPNOTSUPP;
+		goto out;
+	}
+
+	device->default_expires = DIAG_TIMEOUT;
+	device->default_retries = DIAG_MAX_RETRIES;
+
+	/* Figure out position of label block */
+	switch (private->rdc_data.vdev_class) {
+	case DEV_CLASS_FBA:
+		private->pt_block = 1;
+		break;
+	case DEV_CLASS_ECKD:
+		private->pt_block = 2;
+		break;
+	default:
+		pr_warning("%s: Device type %d is not supported "
+			   "in DIAG mode\n", dev_name(&device->cdev->dev),
+			   private->rdc_data.vdev_class);
+		rc = -EOPNOTSUPP;
+		goto out;
+	}
+
+	DBF_DEV_EVENT(DBF_INFO, device,
+		      "%04X: %04X on real %04X/%02X",
+		      rdc_data->dev_nr,
+		      rdc_data->vdev_type,
+		      rdc_data->rdev_type, rdc_data->rdev_model);
+
+	/* terminate all outstanding operations */
+	mdsk_term_io(device);
+
+	/* figure out blocksize of device */
+	label = (struct vtoc_cms_label *) get_zeroed_page(GFP_KERNEL);
+	if (label == NULL)  {
+		DBF_DEV_EVENT(DBF_WARNING, device, "%s",
+			    "No memory to allocate initialization request");
+		rc = -ENOMEM;
+		goto out;
+	}
+	rc = 0;
+	end_block = 0;
+	/* try all sizes - needed for ECKD devices */
+	for (bsize = 512; bsize <= PAGE_SIZE; bsize <<= 1) {
+		mdsk_init_io(device, bsize, 0, &end_block);
+		memset(&bio, 0, sizeof (struct dasd_diag_bio));
+		bio.type = MDSK_READ_REQ;
+		bio.block_number = private->pt_block + 1;
+		bio.buffer = label;
+		memset(&private->iob, 0, sizeof (struct dasd_diag_rw_io));
+		private->iob.dev_nr = rdc_data->dev_nr;
+		private->iob.key = 0;
+		private->iob.flags = 0;	/* do synchronous io */
+		private->iob.block_count = 1;
+		private->iob.interrupt_params = 0;
+		private->iob.bio_list = &bio;
+		private->iob.flaga = DASD_DIAG_FLAGA_DEFAULT;
+		rc = dia250(&private->iob, RW_BIO);
+		if (rc == 3) {
+			pr_warning("%s: A 64-bit DIAG call failed\n",
+				   dev_name(&device->cdev->dev));
+			rc = -EOPNOTSUPP;
+			goto out_label;
+		}
+		mdsk_term_io(device);
+		if (rc == 0)
+			break;
+	}
+	if (bsize > PAGE_SIZE) {
+		pr_warning("%s: Accessing the DASD failed because of an "
+			   "incorrect format (rc=%d)\n",
+			   dev_name(&device->cdev->dev), rc);
+		rc = -EIO;
+		goto out_label;
+	}
+	/* check for label block */
+	if (memcmp(label->label_id, DASD_DIAG_CMS1,
+		  sizeof(DASD_DIAG_CMS1)) == 0) {
+		/* get formatted blocksize from label block */
+		bsize = (unsigned int) label->block_size;
+		block->blocks = (unsigned long) label->block_count;
+	} else
+		block->blocks = end_block;
+	block->bp_block = bsize;
+	block->s2b_shift = 0;	/* bits to shift 512 to get a block */
+	for (sb = 512; sb < bsize; sb = sb << 1)
+		block->s2b_shift++;
+	rc = mdsk_init_io(device, block->bp_block, 0, NULL);
+	if (rc && (rc != 4)) {
+		pr_warning("%s: DIAG initialization failed with rc=%d\n",
+			   dev_name(&device->cdev->dev), rc);
+		rc = -EIO;
+	} else {
+		if (rc == 4)
+			set_bit(DASD_FLAG_DEVICE_RO, &device->flags);
+		pr_info("%s: New DASD with %ld byte/block, total size %ld "
+			"KB%s\n", dev_name(&device->cdev->dev),
+			(unsigned long) block->bp_block,
+			(unsigned long) (block->blocks <<
+					 block->s2b_shift) >> 1,
+			(rc == 4) ? ", read-only device" : "");
+		rc = 0;
+	}
+out_label:
+	free_page((long) label);
+out:
+	if (rc) {
+		device->block = NULL;
+		dasd_free_block(block);
+		device->private = NULL;
+		kfree(private);
+	}
+	return rc;
+}
+
+/* Fill in virtual disk geometry for device. Return zero on success, non-zero
+ * otherwise. */
+static int
+dasd_diag_fill_geometry(struct dasd_block *block, struct hd_geometry *geo)
+{
+	if (dasd_check_blocksize(block->bp_block) != 0)
+		return -EINVAL;
+	geo->cylinders = (block->blocks << block->s2b_shift) >> 10;
+	geo->heads = 16;
+	geo->sectors = 128 >> block->s2b_shift;
+	return 0;
+}
+
+static dasd_erp_fn_t
+dasd_diag_erp_action(struct dasd_ccw_req * cqr)
+{
+	return dasd_default_erp_action;
+}
+
+static dasd_erp_fn_t
+dasd_diag_erp_postaction(struct dasd_ccw_req * cqr)
+{
+	return dasd_default_erp_postaction;
+}
+
+/* Create DASD request from block device request. Return pointer to new
+ * request on success, ERR_PTR otherwise. */
+static struct dasd_ccw_req *dasd_diag_build_cp(struct dasd_device *memdev,
+					       struct dasd_block *block,
+					       struct request *req)
+{
+	struct dasd_ccw_req *cqr;
+	struct dasd_diag_req *dreq;
+	struct dasd_diag_bio *dbio;
+	struct req_iterator iter;
+	struct bio_vec bv;
+	char *dst;
+	unsigned int count, datasize;
+	sector_t recid, first_rec, last_rec;
+	unsigned int blksize, off;
+	unsigned char rw_cmd;
+
+	if (rq_data_dir(req) == READ)
+		rw_cmd = MDSK_READ_REQ;
+	else if (rq_data_dir(req) == WRITE)
+		rw_cmd = MDSK_WRITE_REQ;
+	else
+		return ERR_PTR(-EINVAL);
+	blksize = block->bp_block;
+	/* Calculate record id of first and last block. */
+	first_rec = blk_rq_pos(req) >> block->s2b_shift;
+	last_rec =
+		(blk_rq_pos(req) + blk_rq_sectors(req) - 1) >> block->s2b_shift;
+	/* Check struct bio and count the number of blocks for the request. */
+	count = 0;
+	rq_for_each_segment(bv, req, iter) {
+		if (bv.bv_len & (blksize - 1))
+			/* Fba can only do full blocks. */
+			return ERR_PTR(-EINVAL);
+		count += bv.bv_len >> (block->s2b_shift + 9);
+	}
+	/* Paranoia. */
+	if (count != last_rec - first_rec + 1)
+		return ERR_PTR(-EINVAL);
+	/* Build the request */
+	datasize = sizeof(struct dasd_diag_req) +
+		count*sizeof(struct dasd_diag_bio);
+	cqr = dasd_smalloc_request(DASD_DIAG_MAGIC, 0, datasize, memdev);
+	if (IS_ERR(cqr))
+		return cqr;
+
+	dreq = (struct dasd_diag_req *) cqr->data;
+	dreq->block_count = count;
+	dbio = dreq->bio;
+	recid = first_rec;
+	rq_for_each_segment(bv, req, iter) {
+		dst = page_address(bv.bv_page) + bv.bv_offset;
+		for (off = 0; off < bv.bv_len; off += blksize) {
+			memset(dbio, 0, sizeof (struct dasd_diag_bio));
+			dbio->type = rw_cmd;
+			dbio->block_number = recid + 1;
+			dbio->buffer = dst;
+			dbio++;
+			dst += blksize;
+			recid++;
+		}
+	}
+	cqr->retries = memdev->default_retries;
+	cqr->buildclk = get_tod_clock();
+	if (blk_noretry_request(req) ||
+	    block->base->features & DASD_FEATURE_FAILFAST)
+		set_bit(DASD_CQR_FLAGS_FAILFAST, &cqr->flags);
+	cqr->startdev = memdev;
+	cqr->memdev = memdev;
+	cqr->block = block;
+	cqr->expires = memdev->default_expires * HZ;
+	cqr->status = DASD_CQR_FILLED;
+	return cqr;
+}
+
+/* Release DASD request. Return non-zero if request was successful, zero
+ * otherwise. */
+static int
+dasd_diag_free_cp(struct dasd_ccw_req *cqr, struct request *req)
+{
+	int status;
+
+	status = cqr->status == DASD_CQR_DONE;
+	dasd_sfree_request(cqr, cqr->memdev);
+	return status;
+}
+
+static void dasd_diag_handle_terminated_request(struct dasd_ccw_req *cqr)
+{
+	if (cqr->retries < 0)
+		cqr->status = DASD_CQR_FAILED;
+	else
+		cqr->status = DASD_CQR_FILLED;
+};
+
+/* Fill in IOCTL data for device. */
+static int
+dasd_diag_fill_info(struct dasd_device * device,
+		    struct dasd_information2_t * info)
+{
+	struct dasd_diag_private *private;
+
+	private = (struct dasd_diag_private *) device->private;
+	info->label_block = (unsigned int) private->pt_block;
+	info->FBA_layout = 1;
+	info->format = DASD_FORMAT_LDL;
+	info->characteristics_size = sizeof (struct dasd_diag_characteristics);
+	memcpy(info->characteristics,
+	       &((struct dasd_diag_private *) device->private)->rdc_data,
+	       sizeof (struct dasd_diag_characteristics));
+	info->confdata_size = 0;
+	return 0;
+}
+
+static void
+dasd_diag_dump_sense(struct dasd_device *device, struct dasd_ccw_req * req,
+		     struct irb *stat)
+{
+	DBF_DEV_EVENT(DBF_WARNING, device, "%s",
+		    "dump sense not available for DIAG data");
+}
+
+static struct dasd_discipline dasd_diag_discipline = {
+	.owner = THIS_MODULE,
+	.name = "DIAG",
+	.ebcname = "DIAG",
+	.max_blocks = DIAG_MAX_BLOCKS,
+	.check_device = dasd_diag_check_device,
+	.verify_path = dasd_generic_verify_path,
+	.fill_geometry = dasd_diag_fill_geometry,
+	.start_IO = dasd_start_diag,
+	.term_IO = dasd_diag_term_IO,
+	.handle_terminated_request = dasd_diag_handle_terminated_request,
+	.erp_action = dasd_diag_erp_action,
+	.erp_postaction = dasd_diag_erp_postaction,
+	.build_cp = dasd_diag_build_cp,
+	.free_cp = dasd_diag_free_cp,
+	.dump_sense = dasd_diag_dump_sense,
+	.fill_info = dasd_diag_fill_info,
+};
+
+static int __init
+dasd_diag_init(void)
+{
+	if (!MACHINE_IS_VM) {
+		pr_info("Discipline %s cannot be used without z/VM\n",
+			dasd_diag_discipline.name);
+		return -ENODEV;
+	}
+	ASCEBC(dasd_diag_discipline.ebcname, 4);
+
+	irq_subclass_register(IRQ_SUBCLASS_SERVICE_SIGNAL);
+	register_external_irq(EXT_IRQ_CP_SERVICE, dasd_ext_handler);
+	dasd_diag_discipline_pointer = &dasd_diag_discipline;
+	return 0;
+}
+
+static void __exit
+dasd_diag_cleanup(void)
+{
+	unregister_external_irq(EXT_IRQ_CP_SERVICE, dasd_ext_handler);
+	irq_subclass_unregister(IRQ_SUBCLASS_SERVICE_SIGNAL);
+	dasd_diag_discipline_pointer = NULL;
+}
+
+module_init(dasd_diag_init);
+module_exit(dasd_diag_cleanup);

drivers/s390/block/dasd_diag.h

@@ -0,0 +1,122 @@
+/*
+ * Author(s)......: Holger Smolinski <Holger.Smolinski@de.ibm.com>
+ * Based on.......: linux/drivers/s390/block/mdisk.h
+ * ...............: by Hartmunt Penner <hpenner@de.ibm.com>
+ * Bugreports.to..: <Linux390@de.ibm.com>
+ * Copyright IBM Corp. 1999, 2000
+ *
+ */
+
+#define MDSK_WRITE_REQ 0x01
+#define MDSK_READ_REQ  0x02
+
+#define INIT_BIO	0x00
+#define RW_BIO		0x01
+#define TERM_BIO	0x02
+
+#define DEV_CLASS_FBA	0x01
+#define DEV_CLASS_ECKD	0x04
+
+#define DASD_DIAG_CODE_31BIT		0x03
+#define DASD_DIAG_CODE_64BIT		0x07
+
+#define DASD_DIAG_RWFLAG_ASYNC		0x02
+#define DASD_DIAG_RWFLAG_NOCACHE	0x01
+
+#define DASD_DIAG_FLAGA_FORMAT_64BIT	0x80
+
+struct dasd_diag_characteristics {
+	u16 dev_nr;
+	u16 rdc_len;
+	u8 vdev_class;
+	u8 vdev_type;
+	u8 vdev_status;
+	u8 vdev_flags;
+	u8 rdev_class;
+	u8 rdev_type;
+	u8 rdev_model;
+	u8 rdev_features;
+} __attribute__ ((packed, aligned(4)));
+
+
+#ifdef CONFIG_64BIT
+#define DASD_DIAG_FLAGA_DEFAULT		DASD_DIAG_FLAGA_FORMAT_64BIT
+
+typedef u64 blocknum_t;
+typedef s64 sblocknum_t;
+
+struct dasd_diag_bio {
+	u8 type;
+	u8 status;
+	u8 spare1[2];
+	u32 alet;
+	blocknum_t block_number;
+	void *buffer;
+} __attribute__ ((packed, aligned(8)));
+
+struct dasd_diag_init_io {
+	u16 dev_nr;
+	u8 flaga;
+	u8 spare1[21];
+	u32 block_size;
+	u8 spare2[4];
+	blocknum_t offset;
+	sblocknum_t start_block;
+	blocknum_t end_block;
+	u8  spare3[8];
+} __attribute__ ((packed, aligned(8)));
+
+struct dasd_diag_rw_io {
+	u16 dev_nr;
+	u8  flaga;
+	u8  spare1[21];
+	u8  key;
+	u8  flags;
+	u8  spare2[2];
+	u32 block_count;
+	u32 alet;
+	u8  spare3[4];
+	u64 interrupt_params;
+	struct dasd_diag_bio *bio_list;
+	u8  spare4[8];
+} __attribute__ ((packed, aligned(8)));
+#else /* CONFIG_64BIT */
+#define DASD_DIAG_FLAGA_DEFAULT		0x0
+
+typedef u32 blocknum_t;
+typedef s32 sblocknum_t;
+
+struct dasd_diag_bio {
+	u8 type;
+	u8 status;
+	u16 spare1;
+	blocknum_t block_number;
+	u32 alet;
+	void *buffer;
+} __attribute__ ((packed, aligned(8)));
+
+struct dasd_diag_init_io {
+	u16 dev_nr;
+	u8 flaga;
+	u8 spare1[21];
+	u32 block_size;
+	blocknum_t offset;
+	sblocknum_t start_block;
+	blocknum_t end_block;
+	u8 spare2[24];
+} __attribute__ ((packed, aligned(8)));
+
+struct dasd_diag_rw_io {
+	u16 dev_nr;
+	u8 flaga;
+	u8 spare1[21];
+	u8 key;
+	u8 flags;
+	u8 spare2[2];
+	u32 block_count;
+	u32 alet;
+	struct dasd_diag_bio *bio_list;
+	u32 interrupt_params;
+	u8 spare3[20];
+} __attribute__ ((packed, aligned(8)));
+#endif /* CONFIG_64BIT */

drivers/s390/block/dasd_eckd.h

@@ -0,0 +1,533 @@
+/*
+ * Author(s)......: Holger Smolinski <Holger.Smolinski@de.ibm.com>
+ *		    Horst Hummel <Horst.Hummel@de.ibm.com>
+ * Bugreports.to..: <Linux390@de.ibm.com>
+ * Copyright IBM Corp. 1999, 2000
+ *
+ */
+
+#ifndef DASD_ECKD_H
+#define DASD_ECKD_H
+
+/*****************************************************************************
+ * SECTION: CCW Definitions
+ ****************************************************************************/
+#define DASD_ECKD_CCW_WRITE		 0x05
+#define DASD_ECKD_CCW_READ		 0x06
+#define DASD_ECKD_CCW_WRITE_HOME_ADDRESS 0x09
+#define DASD_ECKD_CCW_READ_HOME_ADDRESS	 0x0a
+#define DASD_ECKD_CCW_WRITE_KD		 0x0d
+#define DASD_ECKD_CCW_READ_KD		 0x0e
+#define DASD_ECKD_CCW_ERASE		 0x11
+#define DASD_ECKD_CCW_READ_COUNT	 0x12
+#define DASD_ECKD_CCW_SLCK		 0x14
+#define DASD_ECKD_CCW_WRITE_RECORD_ZERO	 0x15
+#define DASD_ECKD_CCW_READ_RECORD_ZERO	 0x16
+#define DASD_ECKD_CCW_WRITE_CKD		 0x1d
+#define DASD_ECKD_CCW_READ_CKD		 0x1e
+#define DASD_ECKD_CCW_PSF		 0x27
+#define DASD_ECKD_CCW_SNID		 0x34
+#define DASD_ECKD_CCW_RSSD		 0x3e
+#define DASD_ECKD_CCW_LOCATE_RECORD	 0x47
+#define DASD_ECKD_CCW_SNSS		 0x54
+#define DASD_ECKD_CCW_DEFINE_EXTENT	 0x63
+#define DASD_ECKD_CCW_WRITE_MT		 0x85
+#define DASD_ECKD_CCW_READ_MT		 0x86
+#define DASD_ECKD_CCW_WRITE_KD_MT	 0x8d
+#define DASD_ECKD_CCW_READ_KD_MT	 0x8e
+#define DASD_ECKD_CCW_RELEASE		 0x94
+#define DASD_ECKD_CCW_WRITE_FULL_TRACK	 0x95
+#define DASD_ECKD_CCW_READ_CKD_MT	 0x9e
+#define DASD_ECKD_CCW_WRITE_CKD_MT	 0x9d
+#define DASD_ECKD_CCW_WRITE_TRACK_DATA	 0xA5
+#define DASD_ECKD_CCW_READ_TRACK_DATA	 0xA6
+#define DASD_ECKD_CCW_RESERVE		 0xB4
+#define DASD_ECKD_CCW_READ_TRACK	 0xDE
+#define DASD_ECKD_CCW_PFX		 0xE7
+#define DASD_ECKD_CCW_PFX_READ		 0xEA
+#define DASD_ECKD_CCW_RSCK		 0xF9
+#define DASD_ECKD_CCW_RCD		 0xFA
+
+/*
+ * Perform Subsystem Function / Sub-Orders
+ */
+#define PSF_ORDER_PRSSD			 0x18
+#define PSF_ORDER_CUIR_RESPONSE		 0x1A
+#define PSF_ORDER_SSC			 0x1D
+
+/*
+ * CUIR response condition codes
+ */
+#define PSF_CUIR_INVALID		 0x00
+#define PSF_CUIR_COMPLETED		 0x01
+#define PSF_CUIR_NOT_SUPPORTED		 0x02
+#define PSF_CUIR_ERROR_IN_REQ		 0x03
+#define PSF_CUIR_DENIED			 0x04
+#define PSF_CUIR_LAST_PATH		 0x05
+#define PSF_CUIR_DEVICE_ONLINE		 0x06
+#define PSF_CUIR_VARY_FAILURE		 0x07
+#define PSF_CUIR_SOFTWARE_FAILURE	 0x08
+#define PSF_CUIR_NOT_RECOGNIZED		 0x09
+
+/*
+ * CUIR codes
+ */
+#define CUIR_QUIESCE			 0x01
+#define CUIR_RESUME			 0x02
+
+/*
+ * attention message definitions
+ */
+#define ATTENTION_LENGTH_CUIR		 0x0e
+#define ATTENTION_FORMAT_CUIR		 0x01
+
+/*
+ * Size that is reportet for large volumes in the old 16-bit no_cyl field
+ */
+#define LV_COMPAT_CYL 0xFFFE
+
+
+#define FCX_MAX_DATA_FACTOR 65536
+#define DASD_ECKD_RCD_DATA_SIZE 256
+
+
+/*****************************************************************************
+ * SECTION: Type Definitions
+ ****************************************************************************/
+
+struct eckd_count {
+	__u16 cyl;
+	__u16 head;
+	__u8 record;
+	__u8 kl;
+	__u16 dl;
+} __attribute__ ((packed));
+
+struct ch_t {
+	__u16 cyl;
+	__u16 head;
+} __attribute__ ((packed));
+
+struct chs_t {
+	__u16 cyl;
+	__u16 head;
+	__u32 sector;
+} __attribute__ ((packed));
+
+struct chr_t {
+	__u16 cyl;
+	__u16 head;
+	__u8 record;
+} __attribute__ ((packed));
+
+struct geom_t {
+	__u16 cyl;
+	__u16 head;
+	__u32 sector;
+} __attribute__ ((packed));
+
+struct eckd_home {
+	__u8 skip_control[14];
+	__u16 cell_number;
+	__u8 physical_addr[3];
+	__u8 flag;
+	struct ch_t track_addr;
+	__u8 reserved;
+	__u8 key_length;
+	__u8 reserved2[2];
+} __attribute__ ((packed));
+
+struct DE_eckd_data {
+	struct {
+		unsigned char perm:2;	/* Permissions on this extent */
+		unsigned char reserved:1;
+		unsigned char seek:2;	/* Seek control */
+		unsigned char auth:2;	/* Access authorization */
+		unsigned char pci:1;	/* PCI Fetch mode */
+	} __attribute__ ((packed)) mask;
+	struct {
+		unsigned char mode:2;	/* Architecture mode */
+		unsigned char ckd:1;	/* CKD Conversion */
+		unsigned char operation:3;	/* Operation mode */
+		unsigned char cfw:1;	/* Cache fast write */
+		unsigned char dfw:1;	/* DASD fast write */
+	} __attribute__ ((packed)) attributes;
+	__u16 blk_size;		/* Blocksize */
+	__u16 fast_write_id;
+	__u8 ga_additional;	/* Global Attributes Additional */
+	__u8 ga_extended;	/* Global Attributes Extended	*/
+	struct ch_t beg_ext;
+	struct ch_t end_ext;
+	unsigned long long ep_sys_time; /* Ext Parameter - System Time Stamp */
+	__u8 ep_format;        /* Extended Parameter format byte       */
+	__u8 ep_prio;          /* Extended Parameter priority I/O byte */
+	__u8 ep_reserved1;     /* Extended Parameter Reserved	       */
+	__u8 ep_rec_per_track; /* Number of records on a track	       */
+	__u8 ep_reserved[4];   /* Extended Parameter Reserved	       */
+} __attribute__ ((packed));
+
+struct LO_eckd_data {
+	struct {
+		unsigned char orientation:2;
+		unsigned char operation:6;
+	} __attribute__ ((packed)) operation;
+	struct {
+		unsigned char last_bytes_used:1;
+		unsigned char reserved:6;
+		unsigned char read_count_suffix:1;
+	} __attribute__ ((packed)) auxiliary;
+	__u8 unused;
+	__u8 count;
+	struct ch_t seek_addr;
+	struct chr_t search_arg;
+	__u8 sector;
+	__u16 length;
+} __attribute__ ((packed));
+
+struct LRE_eckd_data {
+	struct {
+		unsigned char orientation:2;
+		unsigned char operation:6;
+	} __attribute__ ((packed)) operation;
+	struct {
+		unsigned char length_valid:1;
+		unsigned char length_scope:1;
+		unsigned char imbedded_ccw_valid:1;
+		unsigned char check_bytes:2;
+		unsigned char imbedded_count_valid:1;
+		unsigned char reserved:1;
+		unsigned char read_count_suffix:1;
+	} __attribute__ ((packed)) auxiliary;
+	__u8 imbedded_ccw;
+	__u8 count;
+	struct ch_t seek_addr;
+	struct chr_t search_arg;
+	__u8 sector;
+	__u16 length;
+	__u8 imbedded_count;
+	__u8 extended_operation;
+	__u16 extended_parameter_length;
+	__u8 extended_parameter[0];
+} __attribute__ ((packed));
+
+/* Prefix data for format 0x00 and 0x01 */
+struct PFX_eckd_data {
+	unsigned char format;
+	struct {
+		unsigned char define_extent:1;
+		unsigned char time_stamp:1;
+		unsigned char verify_base:1;
+		unsigned char hyper_pav:1;
+		unsigned char reserved:4;
+	} __attribute__ ((packed)) validity;
+	__u8 base_address;
+	__u8 aux;
+	__u8 base_lss;
+	__u8 reserved[7];
+	struct DE_eckd_data define_extent;
+	struct LRE_eckd_data locate_record;
+} __attribute__ ((packed));
+
+struct dasd_eckd_characteristics {
+	__u16 cu_type;
+	struct {
+		unsigned char support:2;
+		unsigned char async:1;
+		unsigned char reserved:1;
+		unsigned char cache_info:1;
+		unsigned char model:3;
+	} __attribute__ ((packed)) cu_model;
+	__u16 dev_type;
+	__u8 dev_model;
+	struct {
+		unsigned char mult_burst:1;
+		unsigned char RT_in_LR:1;
+		unsigned char reserved1:1;
+		unsigned char RD_IN_LR:1;
+		unsigned char reserved2:4;
+		unsigned char reserved3:8;
+		unsigned char defect_wr:1;
+		unsigned char XRC_supported:1;
+		unsigned char reserved4:1;
+		unsigned char striping:1;
+		unsigned char reserved5:4;
+		unsigned char cfw:1;
+		unsigned char reserved6:2;
+		unsigned char cache:1;
+		unsigned char dual_copy:1;
+		unsigned char dfw:1;
+		unsigned char reset_alleg:1;
+		unsigned char sense_down:1;
+	} __attribute__ ((packed)) facilities;
+	__u8 dev_class;
+	__u8 unit_type;
+	__u16 no_cyl;
+	__u16 trk_per_cyl;
+	__u8 sec_per_trk;
+	__u8 byte_per_track[3];
+	__u16 home_bytes;
+	__u8 formula;
+	union {
+		struct {
+			__u8 f1;
+			__u16 f2;
+			__u16 f3;
+		} __attribute__ ((packed)) f_0x01;
+		struct {
+			__u8 f1;
+			__u8 f2;
+			__u8 f3;
+			__u8 f4;
+			__u8 f5;
+		} __attribute__ ((packed)) f_0x02;
+	} __attribute__ ((packed)) factors;
+	__u16 first_alt_trk;
+	__u16 no_alt_trk;
+	__u16 first_dia_trk;
+	__u16 no_dia_trk;
+	__u16 first_sup_trk;
+	__u16 no_sup_trk;
+	__u8 MDR_ID;
+	__u8 OBR_ID;
+	__u8 director;
+	__u8 rd_trk_set;
+	__u16 max_rec_zero;
+	__u8 reserved1;
+	__u8 RWANY_in_LR;
+	__u8 factor6;
+	__u8 factor7;
+	__u8 factor8;
+	__u8 reserved2[3];
+	__u8 reserved3[6];
+	__u32 long_no_cyl;
+} __attribute__ ((packed));
+
+/* elements of the configuration data */
+struct dasd_ned {
+	struct {
+		__u8 identifier:2;
+		__u8 token_id:1;
+		__u8 sno_valid:1;
+		__u8 subst_sno:1;
+		__u8 recNED:1;
+		__u8 emuNED:1;
+		__u8 reserved:1;
+	} __attribute__ ((packed)) flags;
+	__u8 descriptor;
+	__u8 dev_class;
+	__u8 reserved;
+	__u8 dev_type[6];
+	__u8 dev_model[3];
+	__u8 HDA_manufacturer[3];
+	__u8 HDA_location[2];
+	__u8 HDA_seqno[12];
+	__u8 ID;
+	__u8 unit_addr;
+} __attribute__ ((packed));
+
+struct dasd_sneq {
+	struct {
+		__u8 identifier:2;
+		__u8 reserved:6;
+	} __attribute__ ((packed)) flags;
+	__u8 res1;
+	__u16 format;
+	__u8 res2[4];		/* byte  4- 7 */
+	__u8 sua_flags;		/* byte  8    */
+	__u8 base_unit_addr;	/* byte  9    */
+	__u8 res3[22];		/* byte 10-31 */
+} __attribute__ ((packed));
+
+struct vd_sneq {
+	struct {
+		__u8 identifier:2;
+		__u8 reserved:6;
+	} __attribute__ ((packed)) flags;
+	__u8 res1;
+	__u16 format;
+	__u8 res2[4];	/* byte  4- 7 */
+	__u8 uit[16];	/* byte  8-23 */
+	__u8 res3[8];	/* byte 24-31 */
+} __attribute__ ((packed));
+
+struct dasd_gneq {
+	struct {
+		__u8 identifier:2;
+		__u8 reserved:6;
+	} __attribute__ ((packed)) flags;
+	__u8 reserved[5];
+	struct {
+		__u8 value:2;
+		__u8 number:6;
+	} __attribute__ ((packed)) timeout;
+	__u8 reserved3;
+	__u16 subsystemID;
+	__u8 reserved2[22];
+} __attribute__ ((packed));
+
+struct dasd_rssd_features {
+	char feature[256];
+} __attribute__((packed));
+
+struct dasd_rssd_messages {
+	__u16 length;
+	__u8 format;
+	__u8 code;
+	__u32 message_id;
+	__u8 flags;
+	char messages[4087];
+} __packed;
+
+struct dasd_cuir_message {
+	__u16 length;
+	__u8 format;
+	__u8 code;
+	__u32 message_id;
+	__u8 flags;
+	__u8 neq_map[3];
+	__u8 ned_map;
+	__u8 record_selector;
+} __packed;
+
+struct dasd_psf_cuir_response {
+	__u8 order;
+	__u8 flags;
+	__u8 cc;
+	__u8 chpid;
+	__u16 device_nr;
+	__u16 reserved;
+	__u32 message_id;
+	__u64 system_id;
+	__u8 cssid;
+	__u8 ssid;
+} __packed;
+
+/*
+ * Perform Subsystem Function - Prepare for Read Subsystem Data
+ */
+struct dasd_psf_prssd_data {
+	unsigned char order;
+	unsigned char flags;
+	unsigned char reserved[4];
+	unsigned char suborder;
+	unsigned char varies[5];
+} __attribute__ ((packed));
+
+/*
+ * Perform Subsystem Function - Set Subsystem Characteristics
+ */
+struct dasd_psf_ssc_data {
+	unsigned char order;
+	unsigned char flags;
+	unsigned char cu_type[4];
+	unsigned char suborder;
+	unsigned char reserved[59];
+} __attribute__((packed));
+
+
+/*
+ * some structures and definitions for alias handling
+ */
+struct dasd_unit_address_configuration {
+	struct {
+		char ua_type;
+		char base_ua;
+	} unit[256];
+} __attribute__((packed));
+
+
+#define MAX_DEVICES_PER_LCU 256
+
+/* flags on the LCU  */
+#define NEED_UAC_UPDATE  0x01
+#define UPDATE_PENDING	0x02
+
+enum pavtype {NO_PAV, BASE_PAV, HYPER_PAV};
+
+
+struct alias_root {
+	struct list_head serverlist;
+	spinlock_t lock;
+};
+
+struct alias_server {
+	struct list_head server;
+	struct dasd_uid uid;
+	struct list_head lculist;
+};
+
+struct summary_unit_check_work_data {
+	char reason;
+	struct dasd_device *device;
+	struct work_struct worker;
+};
+
+struct read_uac_work_data {
+	struct dasd_device *device;
+	struct delayed_work dwork;
+};
+
+struct alias_lcu {
+	struct list_head lcu;
+	struct dasd_uid uid;
+	enum pavtype pav;
+	char flags;
+	spinlock_t lock;
+	struct list_head grouplist;
+	struct list_head active_devices;
+	struct list_head inactive_devices;
+	struct dasd_unit_address_configuration *uac;
+	struct summary_unit_check_work_data suc_data;
+	struct read_uac_work_data ruac_data;
+	struct dasd_ccw_req *rsu_cqr;
+	struct completion lcu_setup;
+};
+
+struct alias_pav_group {
+	struct list_head group;
+	struct dasd_uid uid;
+	struct alias_lcu *lcu;
+	struct list_head baselist;
+	struct list_head aliaslist;
+	struct dasd_device *next;
+};
+
+struct dasd_eckd_private {
+	struct dasd_eckd_characteristics rdc_data;
+	u8 *conf_data;
+	int conf_len;
+	/* pointers to specific parts in the conf_data */
+	struct dasd_ned *ned;
+	struct dasd_sneq *sneq;
+	struct vd_sneq *vdsneq;
+	struct dasd_gneq *gneq;
+
+	struct eckd_count count_area[5];
+	int init_cqr_status;
+	int uses_cdl;
+	struct attrib_data_t attrib;	/* e.g. cache operations */
+	struct dasd_rssd_features features;
+	u32 real_cyl;
+
+	/* alias managemnet */
+	struct dasd_uid uid;
+	struct alias_pav_group *pavgroup;
+	struct alias_lcu *lcu;
+	int count;
+
+	u32 fcx_max_data;
+};
+
+
+
+int dasd_alias_make_device_known_to_lcu(struct dasd_device *);
+void dasd_alias_disconnect_device_from_lcu(struct dasd_device *);
+int dasd_alias_add_device(struct dasd_device *);
+int dasd_alias_remove_device(struct dasd_device *);
+struct dasd_device *dasd_alias_get_start_dev(struct dasd_device *);
+void dasd_alias_handle_summary_unit_check(struct dasd_device *, struct irb *);
+void dasd_eckd_reset_ccw_to_base_io(struct dasd_ccw_req *);
+void dasd_alias_lcu_setup_complete(struct dasd_device *);
+void dasd_alias_wait_for_lcu_setup(struct dasd_device *);
+int dasd_alias_update_add_device(struct dasd_device *);
+#endif				/* DASD_ECKD_H */

drivers/s390/block/dasd_eer.c

@@ -0,0 +1,709 @@
+/*
+ *  Character device driver for extended error reporting.
+ *
+ *  Copyright IBM Corp. 2005
+ *  extended error reporting for DASD ECKD devices
+ *  Author(s): Stefan Weinhuber <wein@de.ibm.com>
+ */
+
+#define KMSG_COMPONENT "dasd-eckd"
+
+#include <linux/init.h>
+#include <linux/fs.h>
+#include <linux/kernel.h>
+#include <linux/miscdevice.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/device.h>
+#include <linux/poll.h>
+#include <linux/mutex.h>
+#include <linux/err.h>
+#include <linux/slab.h>
+
+#include <asm/uaccess.h>
+#include <linux/atomic.h>
+#include <asm/ebcdic.h>
+
+#include "dasd_int.h"
+#include "dasd_eckd.h"
+
+#ifdef PRINTK_HEADER
+#undef PRINTK_HEADER
+#endif				/* PRINTK_HEADER */
+#define PRINTK_HEADER "dasd(eer):"
+
+/*
+ * SECTION: the internal buffer
+ */
+
+/*
+ * The internal buffer is meant to store obaque blobs of data, so it does
+ * not know of higher level concepts like triggers.
+ * It consists of a number of pages that are used as a ringbuffer. Each data
+ * blob is stored in a simple record that consists of an integer, which
+ * contains the size of the following data, and the data bytes themselfes.
+ *
+ * To allow for multiple independent readers we create one internal buffer
+ * each time the device is opened and destroy the buffer when the file is
+ * closed again. The number of pages used for this buffer is determined by
+ * the module parmeter eer_pages.
+ *
+ * One record can be written to a buffer by using the functions
+ * - dasd_eer_start_record (one time per record to write the size to the
+ *                          buffer and reserve the space for the data)
+ * - dasd_eer_write_buffer (one or more times per record to write the data)
+ * The data can be written in several steps but you will have to compute
+ * the total size up front for the invocation of dasd_eer_start_record.
+ * If the ringbuffer is full, dasd_eer_start_record will remove the required
+ * number of old records.
+ *
+ * A record is typically read in two steps, first read the integer that
+ * specifies the size of the following data, then read the data.
+ * Both can be done by
+ * - dasd_eer_read_buffer
+ *
+ * For all mentioned functions you need to get the bufferlock first and keep
+ * it until a complete record is written or read.
+ *
+ * All information necessary to keep track of an internal buffer is kept in
+ * a struct eerbuffer. The buffer specific to a file pointer is strored in
+ * the private_data field of that file. To be able to write data to all
+ * existing buffers, each buffer is also added to the bufferlist.
+ * If the user does not want to read a complete record in one go, we have to
+ * keep track of the rest of the record. residual stores the number of bytes
+ * that are still to deliver. If the rest of the record is invalidated between
+ * two reads then residual will be set to -1 so that the next read will fail.
+ * All entries in the eerbuffer structure are protected with the bufferlock.
+ * To avoid races between writing to a buffer on the one side and creating
+ * and destroying buffers on the other side, the bufferlock must also be used
+ * to protect the bufferlist.
+ */
+
+static int eer_pages = 5;
+module_param(eer_pages, int, S_IRUGO|S_IWUSR);
+
+struct eerbuffer {
+	struct list_head list;
+	char **buffer;
+	int buffersize;
+	int buffer_page_count;
+	int head;
+        int tail;
+	int residual;
+};
+
+static LIST_HEAD(bufferlist);
+static DEFINE_SPINLOCK(bufferlock);
+static DECLARE_WAIT_QUEUE_HEAD(dasd_eer_read_wait_queue);
+
+/*
+ * How many free bytes are available on the buffer.
+ * Needs to be called with bufferlock held.
+ */
+static int dasd_eer_get_free_bytes(struct eerbuffer *eerb)
+{
+	if (eerb->head < eerb->tail)
+		return eerb->tail - eerb->head - 1;
+	return eerb->buffersize - eerb->head + eerb->tail -1;
+}
+
+/*
+ * How many bytes of buffer space are used.
+ * Needs to be called with bufferlock held.
+ */
+static int dasd_eer_get_filled_bytes(struct eerbuffer *eerb)
+{
+
+	if (eerb->head >= eerb->tail)
+		return eerb->head - eerb->tail;
+	return eerb->buffersize - eerb->tail + eerb->head;
+}
+
+/*
+ * The dasd_eer_write_buffer function just copies count bytes of data
+ * to the buffer. Make sure to call dasd_eer_start_record first, to
+ * make sure that enough free space is available.
+ * Needs to be called with bufferlock held.
+ */
+static void dasd_eer_write_buffer(struct eerbuffer *eerb,
+				  char *data, int count)
+{
+
+	unsigned long headindex,localhead;
+	unsigned long rest, len;
+	char *nextdata;
+
+	nextdata = data;
+	rest = count;
+	while (rest > 0) {
+ 		headindex = eerb->head / PAGE_SIZE;
+ 		localhead = eerb->head % PAGE_SIZE;
+		len = min(rest, PAGE_SIZE - localhead);
+		memcpy(eerb->buffer[headindex]+localhead, nextdata, len);
+		nextdata += len;
+		rest -= len;
+		eerb->head += len;
+		if (eerb->head == eerb->buffersize)
+			eerb->head = 0; /* wrap around */
+		BUG_ON(eerb->head > eerb->buffersize);
+	}
+}
+
+/*
+ * Needs to be called with bufferlock held.
+ */
+static int dasd_eer_read_buffer(struct eerbuffer *eerb, char *data, int count)
+{
+
+	unsigned long tailindex,localtail;
+	unsigned long rest, len, finalcount;
+	char *nextdata;
+
+	finalcount = min(count, dasd_eer_get_filled_bytes(eerb));
+	nextdata = data;
+	rest = finalcount;
+	while (rest > 0) {
+ 		tailindex = eerb->tail / PAGE_SIZE;
+ 		localtail = eerb->tail % PAGE_SIZE;
+		len = min(rest, PAGE_SIZE - localtail);
+		memcpy(nextdata, eerb->buffer[tailindex] + localtail, len);
+		nextdata += len;
+		rest -= len;
+		eerb->tail += len;
+		if (eerb->tail == eerb->buffersize)
+			eerb->tail = 0; /* wrap around */
+		BUG_ON(eerb->tail > eerb->buffersize);
+	}
+	return finalcount;
+}
+
+/*
+ * Whenever you want to write a blob of data to the internal buffer you
+ * have to start by using this function first. It will write the number
+ * of bytes that will be written to the buffer. If necessary it will remove
+ * old records to make room for the new one.
+ * Needs to be called with bufferlock held.
+ */
+static int dasd_eer_start_record(struct eerbuffer *eerb, int count)
+{
+	int tailcount;
+
+	if (count + sizeof(count) > eerb->buffersize)
+		return -ENOMEM;
+	while (dasd_eer_get_free_bytes(eerb) < count + sizeof(count)) {
+		if (eerb->residual > 0) {
+			eerb->tail += eerb->residual;
+			if (eerb->tail >= eerb->buffersize)
+				eerb->tail -= eerb->buffersize;
+			eerb->residual = -1;
+		}
+		dasd_eer_read_buffer(eerb, (char *) &tailcount,
+				     sizeof(tailcount));
+		eerb->tail += tailcount;
+		if (eerb->tail >= eerb->buffersize)
+			eerb->tail -= eerb->buffersize;
+	}
+	dasd_eer_write_buffer(eerb, (char*) &count, sizeof(count));
+
+	return 0;
+};
+
+/*
+ * Release pages that are not used anymore.
+ */
+static void dasd_eer_free_buffer_pages(char **buf, int no_pages)
+{
+	int i;
+
+	for (i = 0; i < no_pages; i++)
+		free_page((unsigned long) buf[i]);
+}
+
+/*
+ * Allocate a new set of memory pages.
+ */
+static int dasd_eer_allocate_buffer_pages(char **buf, int no_pages)
+{
+	int i;
+
+	for (i = 0; i < no_pages; i++) {
+		buf[i] = (char *) get_zeroed_page(GFP_KERNEL);
+		if (!buf[i]) {
+			dasd_eer_free_buffer_pages(buf, i);
+			return -ENOMEM;
+		}
+	}
+	return 0;
+}
+
+/*
+ * SECTION: The extended error reporting functionality
+ */
+
+/*
+ * When a DASD device driver wants to report an error, it calls the
+ * function dasd_eer_write and gives the respective trigger ID as
+ * parameter. Currently there are four kinds of triggers:
+ *
+ * DASD_EER_FATALERROR:  all kinds of unrecoverable I/O problems
+ * DASD_EER_PPRCSUSPEND: PPRC was suspended
+ * DASD_EER_NOPATH:      There is no path to the device left.
+ * DASD_EER_STATECHANGE: The state of the device has changed.
+ *
+ * For the first three triggers all required information can be supplied by
+ * the caller. For these triggers a record is written by the function
+ * dasd_eer_write_standard_trigger.
+ *
+ * The DASD_EER_STATECHANGE trigger is special since a sense subsystem
+ * status ccw need to be executed to gather the necessary sense data first.
+ * The dasd_eer_snss function will queue the SNSS request and the request
+ * callback will then call dasd_eer_write with the DASD_EER_STATCHANGE
+ * trigger.
+ *
+ * To avoid memory allocations at runtime, the necessary memory is allocated
+ * when the extended error reporting is enabled for a device (by
+ * dasd_eer_probe). There is one sense subsystem status request for each
+ * eer enabled DASD device. The presence of the cqr in device->eer_cqr
+ * indicates that eer is enable for the device. The use of the snss request
+ * is protected by the DASD_FLAG_EER_IN_USE bit. When this flag indicates
+ * that the cqr is currently in use, dasd_eer_snss cannot start a second
+ * request but sets the DASD_FLAG_EER_SNSS flag instead. The callback of
+ * the SNSS request will check the bit and call dasd_eer_snss again.
+ */
+
+#define SNSS_DATA_SIZE 44
+
+#define DASD_EER_BUSID_SIZE 10
+struct dasd_eer_header {
+	__u32 total_size;
+	__u32 trigger;
+	__u64 tv_sec;
+	__u64 tv_usec;
+	char busid[DASD_EER_BUSID_SIZE];
+} __attribute__ ((packed));
+
+/*
+ * The following function can be used for those triggers that have
+ * all necessary data available when the function is called.
+ * If the parameter cqr is not NULL, the chain of requests will be searched
+ * for valid sense data, and all valid sense data sets will be added to
+ * the triggers data.
+ */
+static void dasd_eer_write_standard_trigger(struct dasd_device *device,
+					    struct dasd_ccw_req *cqr,
+					    int trigger)
+{
+	struct dasd_ccw_req *temp_cqr;
+	int data_size;
+	struct timeval tv;
+	struct dasd_eer_header header;
+	unsigned long flags;
+	struct eerbuffer *eerb;
+	char *sense;
+
+	/* go through cqr chain and count the valid sense data sets */
+	data_size = 0;
+	for (temp_cqr = cqr; temp_cqr; temp_cqr = temp_cqr->refers)
+		if (dasd_get_sense(&temp_cqr->irb))
+			data_size += 32;
+
+	header.total_size = sizeof(header) + data_size + 4; /* "EOR" */
+	header.trigger = trigger;
+	do_gettimeofday(&tv);
+	header.tv_sec = tv.tv_sec;
+	header.tv_usec = tv.tv_usec;
+	strncpy(header.busid, dev_name(&device->cdev->dev),
+		DASD_EER_BUSID_SIZE);
+
+	spin_lock_irqsave(&bufferlock, flags);
+	list_for_each_entry(eerb, &bufferlist, list) {
+		dasd_eer_start_record(eerb, header.total_size);
+		dasd_eer_write_buffer(eerb, (char *) &header, sizeof(header));
+		for (temp_cqr = cqr; temp_cqr; temp_cqr = temp_cqr->refers) {
+			sense = dasd_get_sense(&temp_cqr->irb);
+			if (sense)
+				dasd_eer_write_buffer(eerb, sense, 32);
+		}
+		dasd_eer_write_buffer(eerb, "EOR", 4);
+	}
+	spin_unlock_irqrestore(&bufferlock, flags);
+	wake_up_interruptible(&dasd_eer_read_wait_queue);
+}
+
+/*
+ * This function writes a DASD_EER_STATECHANGE trigger.
+ */
+static void dasd_eer_write_snss_trigger(struct dasd_device *device,
+					struct dasd_ccw_req *cqr,
+					int trigger)
+{
+	int data_size;
+	int snss_rc;
+	struct timeval tv;
+	struct dasd_eer_header header;
+	unsigned long flags;
+	struct eerbuffer *eerb;
+
+	snss_rc = (cqr->status == DASD_CQR_DONE) ? 0 : -EIO;
+	if (snss_rc)
+		data_size = 0;
+	else
+		data_size = SNSS_DATA_SIZE;
+
+	header.total_size = sizeof(header) + data_size + 4; /* "EOR" */
+	header.trigger = DASD_EER_STATECHANGE;
+	do_gettimeofday(&tv);
+	header.tv_sec = tv.tv_sec;
+	header.tv_usec = tv.tv_usec;
+	strncpy(header.busid, dev_name(&device->cdev->dev),
+		DASD_EER_BUSID_SIZE);
+
+	spin_lock_irqsave(&bufferlock, flags);
+	list_for_each_entry(eerb, &bufferlist, list) {
+		dasd_eer_start_record(eerb, header.total_size);
+		dasd_eer_write_buffer(eerb, (char *) &header , sizeof(header));
+		if (!snss_rc)
+			dasd_eer_write_buffer(eerb, cqr->data, SNSS_DATA_SIZE);
+		dasd_eer_write_buffer(eerb, "EOR", 4);
+	}
+	spin_unlock_irqrestore(&bufferlock, flags);
+	wake_up_interruptible(&dasd_eer_read_wait_queue);
+}
+
+/*
+ * This function is called for all triggers. It calls the appropriate
+ * function that writes the actual trigger records.
+ */
+void dasd_eer_write(struct dasd_device *device, struct dasd_ccw_req *cqr,
+		    unsigned int id)
+{
+	if (!device->eer_cqr)
+		return;
+	switch (id) {
+	case DASD_EER_FATALERROR:
+	case DASD_EER_PPRCSUSPEND:
+		dasd_eer_write_standard_trigger(device, cqr, id);
+		break;
+	case DASD_EER_NOPATH:
+		dasd_eer_write_standard_trigger(device, NULL, id);
+		break;
+	case DASD_EER_STATECHANGE:
+		dasd_eer_write_snss_trigger(device, cqr, id);
+		break;
+	default: /* unknown trigger, so we write it without any sense data */
+		dasd_eer_write_standard_trigger(device, NULL, id);
+		break;
+	}
+}
+EXPORT_SYMBOL(dasd_eer_write);
+
+/*
+ * Start a sense subsystem status request.
+ * Needs to be called with the device held.
+ */
+void dasd_eer_snss(struct dasd_device *device)
+{
+	struct dasd_ccw_req *cqr;
+
+	cqr = device->eer_cqr;
+	if (!cqr)	/* Device not eer enabled. */
+		return;
+	if (test_and_set_bit(DASD_FLAG_EER_IN_USE, &device->flags)) {
+		/* Sense subsystem status request in use. */
+		set_bit(DASD_FLAG_EER_SNSS, &device->flags);
+		return;
+	}
+	/* cdev is already locked, can't use dasd_add_request_head */
+	clear_bit(DASD_FLAG_EER_SNSS, &device->flags);
+	cqr->status = DASD_CQR_QUEUED;
+	list_add(&cqr->devlist, &device->ccw_queue);
+	dasd_schedule_device_bh(device);
+}
+
+/*
+ * Callback function for use with sense subsystem status request.
+ */
+static void dasd_eer_snss_cb(struct dasd_ccw_req *cqr, void *data)
+{
+	struct dasd_device *device = cqr->startdev;
+	unsigned long flags;
+
+	dasd_eer_write(device, cqr, DASD_EER_STATECHANGE);
+	spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+	if (device->eer_cqr == cqr) {
+		clear_bit(DASD_FLAG_EER_IN_USE, &device->flags);
+		if (test_bit(DASD_FLAG_EER_SNSS, &device->flags))
+			/* Another SNSS has been requested in the meantime. */
+			dasd_eer_snss(device);
+		cqr = NULL;
+	}
+	spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
+	if (cqr)
+		/*
+		 * Extended error recovery has been switched off while
+		 * the SNSS request was running. It could even have
+		 * been switched off and on again in which case there
+		 * is a new ccw in device->eer_cqr. Free the "old"
+		 * snss request now.
+		 */
+		dasd_kfree_request(cqr, device);
+}
+
+/*
+ * Enable error reporting on a given device.
+ */
+int dasd_eer_enable(struct dasd_device *device)
+{
+	struct dasd_ccw_req *cqr;
+	unsigned long flags;
+	struct ccw1 *ccw;
+
+	if (device->eer_cqr)
+		return 0;
+
+	if (!device->discipline || strcmp(device->discipline->name, "ECKD"))
+		return -EPERM;	/* FIXME: -EMEDIUMTYPE ? */
+
+	cqr = dasd_kmalloc_request(DASD_ECKD_MAGIC, 1 /* SNSS */,
+				   SNSS_DATA_SIZE, device);
+	if (IS_ERR(cqr))
+		return -ENOMEM;
+
+	cqr->startdev = device;
+	cqr->retries = 255;
+	cqr->expires = 10 * HZ;
+	clear_bit(DASD_CQR_FLAGS_USE_ERP, &cqr->flags);
+	set_bit(DASD_CQR_ALLOW_SLOCK, &cqr->flags);
+
+	ccw = cqr->cpaddr;
+	ccw->cmd_code = DASD_ECKD_CCW_SNSS;
+	ccw->count = SNSS_DATA_SIZE;
+	ccw->flags = 0;
+	ccw->cda = (__u32)(addr_t) cqr->data;
+
+	cqr->buildclk = get_tod_clock();
+	cqr->status = DASD_CQR_FILLED;
+	cqr->callback = dasd_eer_snss_cb;
+
+	spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+	if (!device->eer_cqr) {
+		device->eer_cqr = cqr;
+		cqr = NULL;
+	}
+	spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
+	if (cqr)
+		dasd_kfree_request(cqr, device);
+	return 0;
+}
+
+/*
+ * Disable error reporting on a given device.
+ */
+void dasd_eer_disable(struct dasd_device *device)
+{
+	struct dasd_ccw_req *cqr;
+	unsigned long flags;
+	int in_use;
+
+	if (!device->eer_cqr)
+		return;
+	spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+	cqr = device->eer_cqr;
+	device->eer_cqr = NULL;
+	clear_bit(DASD_FLAG_EER_SNSS, &device->flags);
+	in_use = test_and_clear_bit(DASD_FLAG_EER_IN_USE, &device->flags);
+	spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
+	if (cqr && !in_use)
+		dasd_kfree_request(cqr, device);
+}
+
+/*
+ * SECTION: the device operations
+ */
+
+/*
+ * On the one side we need a lock to access our internal buffer, on the
+ * other side a copy_to_user can sleep. So we need to copy the data we have
+ * to transfer in a readbuffer, which is protected by the readbuffer_mutex.
+ */
+static char readbuffer[PAGE_SIZE];
+static DEFINE_MUTEX(readbuffer_mutex);
+
+static int dasd_eer_open(struct inode *inp, struct file *filp)
+{
+	struct eerbuffer *eerb;
+	unsigned long flags;
+
+	eerb = kzalloc(sizeof(struct eerbuffer), GFP_KERNEL);
+	if (!eerb)
+		return -ENOMEM;
+	eerb->buffer_page_count = eer_pages;
+	if (eerb->buffer_page_count < 1 ||
+	    eerb->buffer_page_count > INT_MAX / PAGE_SIZE) {
+		kfree(eerb);
+		DBF_EVENT(DBF_WARNING, "can't open device since module "
+			"parameter eer_pages is smaller than 1 or"
+			" bigger than %d", (int)(INT_MAX / PAGE_SIZE));
+		return -EINVAL;
+	}
+	eerb->buffersize = eerb->buffer_page_count * PAGE_SIZE;
+	eerb->buffer = kmalloc(eerb->buffer_page_count * sizeof(char *),
+			       GFP_KERNEL);
+        if (!eerb->buffer) {
+		kfree(eerb);
+                return -ENOMEM;
+	}
+	if (dasd_eer_allocate_buffer_pages(eerb->buffer,
+					   eerb->buffer_page_count)) {
+		kfree(eerb->buffer);
+		kfree(eerb);
+		return -ENOMEM;
+	}
+	filp->private_data = eerb;
+	spin_lock_irqsave(&bufferlock, flags);
+	list_add(&eerb->list, &bufferlist);
+	spin_unlock_irqrestore(&bufferlock, flags);
+
+	return nonseekable_open(inp,filp);
+}
+
+static int dasd_eer_close(struct inode *inp, struct file *filp)
+{
+	struct eerbuffer *eerb;
+	unsigned long flags;
+
+	eerb = (struct eerbuffer *) filp->private_data;
+	spin_lock_irqsave(&bufferlock, flags);
+	list_del(&eerb->list);
+	spin_unlock_irqrestore(&bufferlock, flags);
+	dasd_eer_free_buffer_pages(eerb->buffer, eerb->buffer_page_count);
+	kfree(eerb->buffer);
+	kfree(eerb);
+
+	return 0;
+}
+
+static ssize_t dasd_eer_read(struct file *filp, char __user *buf,
+			     size_t count, loff_t *ppos)
+{
+	int tc,rc;
+	int tailcount,effective_count;
+        unsigned long flags;
+	struct eerbuffer *eerb;
+
+	eerb = (struct eerbuffer *) filp->private_data;
+	if (mutex_lock_interruptible(&readbuffer_mutex))
+		return -ERESTARTSYS;
+
+	spin_lock_irqsave(&bufferlock, flags);
+
+	if (eerb->residual < 0) { /* the remainder of this record */
+		                  /* has been deleted             */
+		eerb->residual = 0;
+		spin_unlock_irqrestore(&bufferlock, flags);
+		mutex_unlock(&readbuffer_mutex);
+		return -EIO;
+	} else if (eerb->residual > 0) {
+		/* OK we still have a second half of a record to deliver */
+		effective_count = min(eerb->residual, (int) count);
+		eerb->residual -= effective_count;
+	} else {
+		tc = 0;
+		while (!tc) {
+			tc = dasd_eer_read_buffer(eerb, (char *) &tailcount,
+						  sizeof(tailcount));
+			if (!tc) {
+				/* no data available */
+				spin_unlock_irqrestore(&bufferlock, flags);
+				mutex_unlock(&readbuffer_mutex);
+				if (filp->f_flags & O_NONBLOCK)
+					return -EAGAIN;
+				rc = wait_event_interruptible(
+					dasd_eer_read_wait_queue,
+					eerb->head != eerb->tail);
+				if (rc)
+					return rc;
+				if (mutex_lock_interruptible(&readbuffer_mutex))
+					return -ERESTARTSYS;
+				spin_lock_irqsave(&bufferlock, flags);
+			}
+		}
+		WARN_ON(tc != sizeof(tailcount));
+		effective_count = min(tailcount,(int)count);
+		eerb->residual = tailcount - effective_count;
+	}
+
+	tc = dasd_eer_read_buffer(eerb, readbuffer, effective_count);
+	WARN_ON(tc != effective_count);
+
+	spin_unlock_irqrestore(&bufferlock, flags);
+
+	if (copy_to_user(buf, readbuffer, effective_count)) {
+		mutex_unlock(&readbuffer_mutex);
+		return -EFAULT;
+	}
+
+	mutex_unlock(&readbuffer_mutex);
+	return effective_count;
+}
+
+static unsigned int dasd_eer_poll(struct file *filp, poll_table *ptable)
+{
+	unsigned int mask;
+	unsigned long flags;
+	struct eerbuffer *eerb;
+
+	eerb = (struct eerbuffer *) filp->private_data;
+	poll_wait(filp, &dasd_eer_read_wait_queue, ptable);
+	spin_lock_irqsave(&bufferlock, flags);
+	if (eerb->head != eerb->tail)
+		mask = POLLIN | POLLRDNORM ;
+	else
+		mask = 0;
+	spin_unlock_irqrestore(&bufferlock, flags);
+	return mask;
+}
+
+static const struct file_operations dasd_eer_fops = {
+	.open		= &dasd_eer_open,
+	.release	= &dasd_eer_close,
+	.read		= &dasd_eer_read,
+	.poll		= &dasd_eer_poll,
+	.owner		= THIS_MODULE,
+	.llseek		= noop_llseek,
+};
+
+static struct miscdevice *dasd_eer_dev = NULL;
+
+int __init dasd_eer_init(void)
+{
+	int rc;
+
+	dasd_eer_dev = kzalloc(sizeof(*dasd_eer_dev), GFP_KERNEL);
+	if (!dasd_eer_dev)
+		return -ENOMEM;
+
+	dasd_eer_dev->minor = MISC_DYNAMIC_MINOR;
+	dasd_eer_dev->name  = "dasd_eer";
+	dasd_eer_dev->fops  = &dasd_eer_fops;
+
+	rc = misc_register(dasd_eer_dev);
+	if (rc) {
+		kfree(dasd_eer_dev);
+		dasd_eer_dev = NULL;
+		DBF_EVENT(DBF_ERR, "%s", "dasd_eer_init could not "
+		       "register misc device");
+		return rc;
+	}
+
+	return 0;
+}
+
+void dasd_eer_exit(void)
+{
+	if (dasd_eer_dev) {
+		misc_deregister(dasd_eer_dev);
+		kfree(dasd_eer_dev);
+		dasd_eer_dev = NULL;
+	}
+}

drivers/s390/block/dasd_erp.c

@@ -0,0 +1,202 @@
+/*
+ * Author(s)......: Holger Smolinski <Holger.Smolinski@de.ibm.com>
+ *		    Horst Hummel <Horst.Hummel@de.ibm.com>
+ *		    Carsten Otte <Cotte@de.ibm.com>
+ *		    Martin Schwidefsky <schwidefsky@de.ibm.com>
+ * Bugreports.to..: <Linux390@de.ibm.com>
+ * Copyright IBM Corp. 1999, 2001
+ *
+ */
+
+#define KMSG_COMPONENT "dasd"
+
+#include <linux/ctype.h>
+#include <linux/init.h>
+
+#include <asm/debug.h>
+#include <asm/ebcdic.h>
+#include <asm/uaccess.h>
+
+/* This is ugly... */
+#define PRINTK_HEADER "dasd_erp:"
+
+#include "dasd_int.h"
+
+struct dasd_ccw_req *
+dasd_alloc_erp_request(char *magic, int cplength, int datasize,
+		       struct dasd_device * device)
+{
+	unsigned long flags;
+	struct dasd_ccw_req *cqr;
+	char *data;
+	int size;
+
+	/* Sanity checks */
+	BUG_ON( magic == NULL || datasize > PAGE_SIZE ||
+	     (cplength*sizeof(struct ccw1)) > PAGE_SIZE);
+
+	size = (sizeof(struct dasd_ccw_req) + 7L) & -8L;
+	if (cplength > 0)
+		size += cplength * sizeof(struct ccw1);
+	if (datasize > 0)
+		size += datasize;
+	spin_lock_irqsave(&device->mem_lock, flags);
+	cqr = (struct dasd_ccw_req *)
+		dasd_alloc_chunk(&device->erp_chunks, size);
+	spin_unlock_irqrestore(&device->mem_lock, flags);
+	if (cqr == NULL)
+		return ERR_PTR(-ENOMEM);
+	memset(cqr, 0, sizeof(struct dasd_ccw_req));
+	INIT_LIST_HEAD(&cqr->devlist);
+	INIT_LIST_HEAD(&cqr->blocklist);
+	data = (char *) cqr + ((sizeof(struct dasd_ccw_req) + 7L) & -8L);
+	cqr->cpaddr = NULL;
+	if (cplength > 0) {
+		cqr->cpaddr = (struct ccw1 *) data;
+		data += cplength*sizeof(struct ccw1);
+		memset(cqr->cpaddr, 0, cplength*sizeof(struct ccw1));
+	}
+	cqr->data = NULL;
+	if (datasize > 0) {
+		cqr->data = data;
+ 		memset(cqr->data, 0, datasize);
+	}
+	strncpy((char *) &cqr->magic, magic, 4);
+	ASCEBC((char *) &cqr->magic, 4);
+	set_bit(DASD_CQR_FLAGS_USE_ERP, &cqr->flags);
+	dasd_get_device(device);
+	return cqr;
+}
+
+void
+dasd_free_erp_request(struct dasd_ccw_req *cqr, struct dasd_device * device)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&device->mem_lock, flags);
+	dasd_free_chunk(&device->erp_chunks, cqr);
+	spin_unlock_irqrestore(&device->mem_lock, flags);
+	atomic_dec(&device->ref_count);
+}
+
+
+/*
+ * dasd_default_erp_action just retries the current cqr
+ */
+struct dasd_ccw_req *
+dasd_default_erp_action(struct dasd_ccw_req *cqr)
+{
+	struct dasd_device *device;
+
+	device = cqr->startdev;
+
+        /* just retry - there is nothing to save ... I got no sense data.... */
+        if (cqr->retries > 0) {
+		DBF_DEV_EVENT(DBF_DEBUG, device,
+                             "default ERP called (%i retries left)",
+                             cqr->retries);
+		if (!test_bit(DASD_CQR_VERIFY_PATH, &cqr->flags))
+			cqr->lpm = device->path_data.opm;
+		cqr->status = DASD_CQR_FILLED;
+        } else {
+		pr_err("%s: default ERP has run out of retries and failed\n",
+		       dev_name(&device->cdev->dev));
+		cqr->status = DASD_CQR_FAILED;
+		cqr->stopclk = get_tod_clock();
+        }
+        return cqr;
+}				/* end dasd_default_erp_action */
+
+/*
+ * DESCRIPTION
+ *   Frees all ERPs of the current ERP Chain and set the status
+ *   of the original CQR either to DASD_CQR_DONE if ERP was successful
+ *   or to DASD_CQR_FAILED if ERP was NOT successful.
+ *   NOTE: This function is only called if no discipline postaction
+ *	   is available
+ *
+ * PARAMETER
+ *   erp		current erp_head
+ *
+ * RETURN VALUES
+ *   cqr		pointer to the original CQR
+ */
+struct dasd_ccw_req *dasd_default_erp_postaction(struct dasd_ccw_req *cqr)
+{
+	int success;
+	unsigned long long startclk, stopclk;
+	struct dasd_device *startdev;
+
+	BUG_ON(cqr->refers == NULL || cqr->function == NULL);
+
+	success = cqr->status == DASD_CQR_DONE;
+	startclk = cqr->startclk;
+	stopclk = cqr->stopclk;
+	startdev = cqr->startdev;
+
+	/* free all ERPs - but NOT the original cqr */
+	while (cqr->refers != NULL) {
+		struct dasd_ccw_req *refers;
+
+		refers = cqr->refers;
+		/* remove the request from the block queue */
+		list_del(&cqr->blocklist);
+		/* free the finished erp request */
+		dasd_free_erp_request(cqr, cqr->memdev);
+		cqr = refers;
+	}
+
+	/* set corresponding status to original cqr */
+	cqr->startclk = startclk;
+	cqr->stopclk = stopclk;
+	cqr->startdev = startdev;
+	if (success)
+		cqr->status = DASD_CQR_DONE;
+	else {
+		cqr->status = DASD_CQR_FAILED;
+		cqr->stopclk = get_tod_clock();
+	}
+
+	return cqr;
+
+}				/* end default_erp_postaction */
+
+void
+dasd_log_sense(struct dasd_ccw_req *cqr, struct irb *irb)
+{
+	struct dasd_device *device;
+
+	device = cqr->startdev;
+	if (cqr->intrc == -ETIMEDOUT) {
+		dev_err(&device->cdev->dev,
+			"A timeout error occurred for cqr %p", cqr);
+		return;
+	}
+	if (cqr->intrc == -ENOLINK) {
+		dev_err(&device->cdev->dev,
+			"A transport error occurred for cqr %p", cqr);
+		return;
+	}
+	/* dump sense data */
+	if (device->discipline && device->discipline->dump_sense)
+		device->discipline->dump_sense(device, cqr, irb);
+}
+
+void
+dasd_log_sense_dbf(struct dasd_ccw_req *cqr, struct irb *irb)
+{
+	struct dasd_device *device;
+
+	device = cqr->startdev;
+	/* dump sense data to s390 debugfeature*/
+	if (device->discipline && device->discipline->dump_sense_dbf)
+		device->discipline->dump_sense_dbf(device, irb, "log");
+}
+EXPORT_SYMBOL(dasd_log_sense_dbf);
+
+EXPORT_SYMBOL(dasd_default_erp_action);
+EXPORT_SYMBOL(dasd_default_erp_postaction);
+EXPORT_SYMBOL(dasd_alloc_erp_request);
+EXPORT_SYMBOL(dasd_free_erp_request);
+EXPORT_SYMBOL(dasd_log_sense);
+

drivers/s390/block/dasd_fba.c

@@ -0,0 +1,635 @@
+/*
+ * Author(s)......: Holger Smolinski <Holger.Smolinski@de.ibm.com>
+ * Bugreports.to..: <Linux390@de.ibm.com>
+ * Copyright IBM Corp. 1999, 2009
+ */
+
+#define KMSG_COMPONENT "dasd-fba"
+
+#include <linux/stddef.h>
+#include <linux/kernel.h>
+#include <asm/debug.h>
+
+#include <linux/slab.h>
+#include <linux/hdreg.h>	/* HDIO_GETGEO			    */
+#include <linux/bio.h>
+#include <linux/module.h>
+#include <linux/init.h>
+
+#include <asm/idals.h>
+#include <asm/ebcdic.h>
+#include <asm/io.h>
+#include <asm/ccwdev.h>
+
+#include "dasd_int.h"
+#include "dasd_fba.h"
+
+#ifdef PRINTK_HEADER
+#undef PRINTK_HEADER
+#endif				/* PRINTK_HEADER */
+#define PRINTK_HEADER "dasd(fba):"
+
+#define FBA_DEFAULT_RETRIES 32
+
+#define DASD_FBA_CCW_WRITE 0x41
+#define DASD_FBA_CCW_READ 0x42
+#define DASD_FBA_CCW_LOCATE 0x43
+#define DASD_FBA_CCW_DEFINE_EXTENT 0x63
+
+MODULE_LICENSE("GPL");
+
+static struct dasd_discipline dasd_fba_discipline;
+
+struct dasd_fba_private {
+	struct dasd_fba_characteristics rdc_data;
+};
+
+static struct ccw_device_id dasd_fba_ids[] = {
+	{ CCW_DEVICE_DEVTYPE (0x6310, 0, 0x9336, 0), .driver_info = 0x1},
+	{ CCW_DEVICE_DEVTYPE (0x3880, 0, 0x3370, 0), .driver_info = 0x2},
+	{ /* end of list */ },
+};
+
+MODULE_DEVICE_TABLE(ccw, dasd_fba_ids);
+
+static struct ccw_driver dasd_fba_driver; /* see below */
+static int
+dasd_fba_probe(struct ccw_device *cdev)
+{
+	return dasd_generic_probe(cdev, &dasd_fba_discipline);
+}
+
+static int
+dasd_fba_set_online(struct ccw_device *cdev)
+{
+	return dasd_generic_set_online(cdev, &dasd_fba_discipline);
+}
+
+static struct ccw_driver dasd_fba_driver = {
+	.driver = {
+		.name	= "dasd-fba",
+		.owner	= THIS_MODULE,
+	},
+	.ids         = dasd_fba_ids,
+	.probe       = dasd_fba_probe,
+	.remove      = dasd_generic_remove,
+	.set_offline = dasd_generic_set_offline,
+	.set_online  = dasd_fba_set_online,
+	.notify      = dasd_generic_notify,
+	.path_event  = dasd_generic_path_event,
+	.freeze      = dasd_generic_pm_freeze,
+	.thaw	     = dasd_generic_restore_device,
+	.restore     = dasd_generic_restore_device,
+	.int_class   = IRQIO_DAS,
+};
+
+static void
+define_extent(struct ccw1 * ccw, struct DE_fba_data *data, int rw,
+	      int blksize, int beg, int nr)
+{
+	ccw->cmd_code = DASD_FBA_CCW_DEFINE_EXTENT;
+	ccw->flags = 0;
+	ccw->count = 16;
+	ccw->cda = (__u32) __pa(data);
+	memset(data, 0, sizeof (struct DE_fba_data));
+	if (rw == WRITE)
+		(data->mask).perm = 0x0;
+	else if (rw == READ)
+		(data->mask).perm = 0x1;
+	else
+		data->mask.perm = 0x2;
+	data->blk_size = blksize;
+	data->ext_loc = beg;
+	data->ext_end = nr - 1;
+}
+
+static void
+locate_record(struct ccw1 * ccw, struct LO_fba_data *data, int rw,
+	      int block_nr, int block_ct)
+{
+	ccw->cmd_code = DASD_FBA_CCW_LOCATE;
+	ccw->flags = 0;
+	ccw->count = 8;
+	ccw->cda = (__u32) __pa(data);
+	memset(data, 0, sizeof (struct LO_fba_data));
+	if (rw == WRITE)
+		data->operation.cmd = 0x5;
+	else if (rw == READ)
+		data->operation.cmd = 0x6;
+	else
+		data->operation.cmd = 0x8;
+	data->blk_nr = block_nr;
+	data->blk_ct = block_ct;
+}
+
+static int
+dasd_fba_check_characteristics(struct dasd_device *device)
+{
+	struct dasd_block *block;
+	struct dasd_fba_private *private;
+	struct ccw_device *cdev = device->cdev;
+	int rc;
+	int readonly;
+
+	private = (struct dasd_fba_private *) device->private;
+	if (!private) {
+		private = kzalloc(sizeof(*private), GFP_KERNEL | GFP_DMA);
+		if (!private) {
+			dev_warn(&device->cdev->dev,
+				 "Allocating memory for private DASD "
+				 "data failed\n");
+			return -ENOMEM;
+		}
+		device->private = (void *) private;
+	} else {
+		memset(private, 0, sizeof(*private));
+	}
+	block = dasd_alloc_block();
+	if (IS_ERR(block)) {
+		DBF_EVENT_DEVID(DBF_WARNING, cdev, "%s", "could not allocate "
+				"dasd block structure");
+		device->private = NULL;
+		kfree(private);
+		return PTR_ERR(block);
+	}
+	device->block = block;
+	block->base = device;
+
+	/* Read Device Characteristics */
+	rc = dasd_generic_read_dev_chars(device, DASD_FBA_MAGIC,
+					 &private->rdc_data, 32);
+	if (rc) {
+		DBF_EVENT_DEVID(DBF_WARNING, cdev, "Read device "
+				"characteristics returned error %d", rc);
+		device->block = NULL;
+		dasd_free_block(block);
+		device->private = NULL;
+		kfree(private);
+		return rc;
+	}
+
+	device->default_expires = DASD_EXPIRES;
+	device->default_retries = FBA_DEFAULT_RETRIES;
+	device->path_data.opm = LPM_ANYPATH;
+
+	readonly = dasd_device_is_ro(device);
+	if (readonly)
+		set_bit(DASD_FLAG_DEVICE_RO, &device->flags);
+
+	dev_info(&device->cdev->dev,
+		 "New FBA DASD %04X/%02X (CU %04X/%02X) with %d MB "
+		 "and %d B/blk%s\n",
+		 cdev->id.dev_type,
+		 cdev->id.dev_model,
+		 cdev->id.cu_type,
+		 cdev->id.cu_model,
+		 ((private->rdc_data.blk_bdsa *
+		   (private->rdc_data.blk_size >> 9)) >> 11),
+		 private->rdc_data.blk_size,
+		 readonly ? ", read-only device" : "");
+	return 0;
+}
+
+static int dasd_fba_do_analysis(struct dasd_block *block)
+{
+	struct dasd_fba_private *private;
+	int sb, rc;
+
+	private = (struct dasd_fba_private *) block->base->private;
+	rc = dasd_check_blocksize(private->rdc_data.blk_size);
+	if (rc) {
+		DBF_DEV_EVENT(DBF_WARNING, block->base, "unknown blocksize %d",
+			    private->rdc_data.blk_size);
+		return rc;
+	}
+	block->blocks = private->rdc_data.blk_bdsa;
+	block->bp_block = private->rdc_data.blk_size;
+	block->s2b_shift = 0;	/* bits to shift 512 to get a block */
+	for (sb = 512; sb < private->rdc_data.blk_size; sb = sb << 1)
+		block->s2b_shift++;
+	return 0;
+}
+
+static int dasd_fba_fill_geometry(struct dasd_block *block,
+				  struct hd_geometry *geo)
+{
+	if (dasd_check_blocksize(block->bp_block) != 0)
+		return -EINVAL;
+	geo->cylinders = (block->blocks << block->s2b_shift) >> 10;
+	geo->heads = 16;
+	geo->sectors = 128 >> block->s2b_shift;
+	return 0;
+}
+
+static dasd_erp_fn_t
+dasd_fba_erp_action(struct dasd_ccw_req * cqr)
+{
+	return dasd_default_erp_action;
+}
+
+static dasd_erp_fn_t
+dasd_fba_erp_postaction(struct dasd_ccw_req * cqr)
+{
+	if (cqr->function == dasd_default_erp_action)
+		return dasd_default_erp_postaction;
+
+	DBF_DEV_EVENT(DBF_WARNING, cqr->startdev, "unknown ERP action %p",
+		    cqr->function);
+	return NULL;
+}
+
+static void dasd_fba_check_for_device_change(struct dasd_device *device,
+					     struct dasd_ccw_req *cqr,
+					     struct irb *irb)
+{
+	char mask;
+
+	/* first of all check for state change pending interrupt */
+	mask = DEV_STAT_ATTENTION | DEV_STAT_DEV_END | DEV_STAT_UNIT_EXCEP;
+	if ((irb->scsw.cmd.dstat & mask) == mask)
+		dasd_generic_handle_state_change(device);
+};
+
+static struct dasd_ccw_req *dasd_fba_build_cp(struct dasd_device * memdev,
+					      struct dasd_block *block,
+					      struct request *req)
+{
+	struct dasd_fba_private *private;
+	unsigned long *idaws;
+	struct LO_fba_data *LO_data;
+	struct dasd_ccw_req *cqr;
+	struct ccw1 *ccw;
+	struct req_iterator iter;
+	struct bio_vec bv;
+	char *dst;
+	int count, cidaw, cplength, datasize;
+	sector_t recid, first_rec, last_rec;
+	unsigned int blksize, off;
+	unsigned char cmd;
+
+	private = (struct dasd_fba_private *) block->base->private;
+	if (rq_data_dir(req) == READ) {
+		cmd = DASD_FBA_CCW_READ;
+	} else if (rq_data_dir(req) == WRITE) {
+		cmd = DASD_FBA_CCW_WRITE;
+	} else
+		return ERR_PTR(-EINVAL);
+	blksize = block->bp_block;
+	/* Calculate record id of first and last block. */
+	first_rec = blk_rq_pos(req) >> block->s2b_shift;
+	last_rec =
+		(blk_rq_pos(req) + blk_rq_sectors(req) - 1) >> block->s2b_shift;
+	/* Check struct bio and count the number of blocks for the request. */
+	count = 0;
+	cidaw = 0;
+	rq_for_each_segment(bv, req, iter) {
+		if (bv.bv_len & (blksize - 1))
+			/* Fba can only do full blocks. */
+			return ERR_PTR(-EINVAL);
+		count += bv.bv_len >> (block->s2b_shift + 9);
+#if defined(CONFIG_64BIT)
+		if (idal_is_needed (page_address(bv.bv_page), bv.bv_len))
+			cidaw += bv.bv_len / blksize;
+#endif
+	}
+	/* Paranoia. */
+	if (count != last_rec - first_rec + 1)
+		return ERR_PTR(-EINVAL);
+	/* 1x define extent + 1x locate record + number of blocks */
+	cplength = 2 + count;
+	/* 1x define extent + 1x locate record */
+	datasize = sizeof(struct DE_fba_data) + sizeof(struct LO_fba_data) +
+		cidaw * sizeof(unsigned long);
+	/*
+	 * Find out number of additional locate record ccws if the device
+	 * can't do data chaining.
+	 */
+	if (private->rdc_data.mode.bits.data_chain == 0) {
+		cplength += count - 1;
+		datasize += (count - 1)*sizeof(struct LO_fba_data);
+	}
+	/* Allocate the ccw request. */
+	cqr = dasd_smalloc_request(DASD_FBA_MAGIC, cplength, datasize, memdev);
+	if (IS_ERR(cqr))
+		return cqr;
+	ccw = cqr->cpaddr;
+	/* First ccw is define extent. */
+	define_extent(ccw++, cqr->data, rq_data_dir(req),
+		      block->bp_block, blk_rq_pos(req), blk_rq_sectors(req));
+	/* Build locate_record + read/write ccws. */
+	idaws = (unsigned long *) (cqr->data + sizeof(struct DE_fba_data));
+	LO_data = (struct LO_fba_data *) (idaws + cidaw);
+	/* Locate record for all blocks for smart devices. */
+	if (private->rdc_data.mode.bits.data_chain != 0) {
+		ccw[-1].flags |= CCW_FLAG_CC;
+		locate_record(ccw++, LO_data++, rq_data_dir(req), 0, count);
+	}
+	recid = first_rec;
+	rq_for_each_segment(bv, req, iter) {
+		dst = page_address(bv.bv_page) + bv.bv_offset;
+		if (dasd_page_cache) {
+			char *copy = kmem_cache_alloc(dasd_page_cache,
+						      GFP_DMA | __GFP_NOWARN);
+			if (copy && rq_data_dir(req) == WRITE)
+				memcpy(copy + bv.bv_offset, dst, bv.bv_len);
+			if (copy)
+				dst = copy + bv.bv_offset;
+		}
+		for (off = 0; off < bv.bv_len; off += blksize) {
+			/* Locate record for stupid devices. */
+			if (private->rdc_data.mode.bits.data_chain == 0) {
+				ccw[-1].flags |= CCW_FLAG_CC;
+				locate_record(ccw, LO_data++,
+					      rq_data_dir(req),
+					      recid - first_rec, 1);
+				ccw->flags = CCW_FLAG_CC;
+				ccw++;
+			} else {
+				if (recid > first_rec)
+					ccw[-1].flags |= CCW_FLAG_DC;
+				else
+					ccw[-1].flags |= CCW_FLAG_CC;
+			}
+			ccw->cmd_code = cmd;
+			ccw->count = block->bp_block;
+			if (idal_is_needed(dst, blksize)) {
+				ccw->cda = (__u32)(addr_t) idaws;
+				ccw->flags = CCW_FLAG_IDA;
+				idaws = idal_create_words(idaws, dst, blksize);
+			} else {
+				ccw->cda = (__u32)(addr_t) dst;
+				ccw->flags = 0;
+			}
+			ccw++;
+			dst += blksize;
+			recid++;
+		}
+	}
+	if (blk_noretry_request(req) ||
+	    block->base->features & DASD_FEATURE_FAILFAST)
+		set_bit(DASD_CQR_FLAGS_FAILFAST, &cqr->flags);
+	cqr->startdev = memdev;
+	cqr->memdev = memdev;
+	cqr->block = block;
+	cqr->expires = memdev->default_expires * HZ;	/* default 5 minutes */
+	cqr->retries = memdev->default_retries;
+	cqr->buildclk = get_tod_clock();
+	cqr->status = DASD_CQR_FILLED;
+	return cqr;
+}
+
+static int
+dasd_fba_free_cp(struct dasd_ccw_req *cqr, struct request *req)
+{
+	struct dasd_fba_private *private;
+	struct ccw1 *ccw;
+	struct req_iterator iter;
+	struct bio_vec bv;
+	char *dst, *cda;
+	unsigned int blksize, off;
+	int status;
+
+	if (!dasd_page_cache)
+		goto out;
+	private = (struct dasd_fba_private *) cqr->block->base->private;
+	blksize = cqr->block->bp_block;
+	ccw = cqr->cpaddr;
+	/* Skip over define extent & locate record. */
+	ccw++;
+	if (private->rdc_data.mode.bits.data_chain != 0)
+		ccw++;
+	rq_for_each_segment(bv, req, iter) {
+		dst = page_address(bv.bv_page) + bv.bv_offset;
+		for (off = 0; off < bv.bv_len; off += blksize) {
+			/* Skip locate record. */
+			if (private->rdc_data.mode.bits.data_chain == 0)
+				ccw++;
+			if (dst) {
+				if (ccw->flags & CCW_FLAG_IDA)
+					cda = *((char **)((addr_t) ccw->cda));
+				else
+					cda = (char *)((addr_t) ccw->cda);
+				if (dst != cda) {
+					if (rq_data_dir(req) == READ)
+						memcpy(dst, cda, bv.bv_len);
+					kmem_cache_free(dasd_page_cache,
+					    (void *)((addr_t)cda & PAGE_MASK));
+				}
+				dst = NULL;
+			}
+			ccw++;
+		}
+	}
+out:
+	status = cqr->status == DASD_CQR_DONE;
+	dasd_sfree_request(cqr, cqr->memdev);
+	return status;
+}
+
+static void dasd_fba_handle_terminated_request(struct dasd_ccw_req *cqr)
+{
+	if (cqr->retries < 0)
+		cqr->status = DASD_CQR_FAILED;
+	else
+		cqr->status = DASD_CQR_FILLED;
+};
+
+static int
+dasd_fba_fill_info(struct dasd_device * device,
+		   struct dasd_information2_t * info)
+{
+	info->label_block = 1;
+	info->FBA_layout = 1;
+	info->format = DASD_FORMAT_LDL;
+	info->characteristics_size = sizeof(struct dasd_fba_characteristics);
+	memcpy(info->characteristics,
+	       &((struct dasd_fba_private *) device->private)->rdc_data,
+	       sizeof (struct dasd_fba_characteristics));
+	info->confdata_size = 0;
+	return 0;
+}
+
+static void
+dasd_fba_dump_sense_dbf(struct dasd_device *device, struct irb *irb,
+			char *reason)
+{
+	u64 *sense;
+
+	sense = (u64 *) dasd_get_sense(irb);
+	if (sense) {
+		DBF_DEV_EVENT(DBF_EMERG, device,
+			      "%s: %s %02x%02x%02x %016llx %016llx %016llx "
+			      "%016llx", reason,
+			      scsw_is_tm(&irb->scsw) ? "t" : "c",
+			      scsw_cc(&irb->scsw), scsw_cstat(&irb->scsw),
+			      scsw_dstat(&irb->scsw), sense[0], sense[1],
+			      sense[2], sense[3]);
+	} else {
+		DBF_DEV_EVENT(DBF_EMERG, device, "%s",
+			      "SORRY - NO VALID SENSE AVAILABLE\n");
+	}
+}
+
+
+static void
+dasd_fba_dump_sense(struct dasd_device *device, struct dasd_ccw_req * req,
+		    struct irb *irb)
+{
+	char *page;
+	struct ccw1 *act, *end, *last;
+	int len, sl, sct, count;
+
+	page = (char *) get_zeroed_page(GFP_ATOMIC);
+	if (page == NULL) {
+		DBF_DEV_EVENT(DBF_WARNING, device, "%s",
+			    "No memory to dump sense data");
+		return;
+	}
+	len = sprintf(page, PRINTK_HEADER
+		      " I/O status report for device %s:\n",
+		      dev_name(&device->cdev->dev));
+	len += sprintf(page + len, PRINTK_HEADER
+		       " in req: %p CS: 0x%02X DS: 0x%02X\n", req,
+		       irb->scsw.cmd.cstat, irb->scsw.cmd.dstat);
+	len += sprintf(page + len, PRINTK_HEADER
+		       " device %s: Failing CCW: %p\n",
+		       dev_name(&device->cdev->dev),
+		       (void *) (addr_t) irb->scsw.cmd.cpa);
+	if (irb->esw.esw0.erw.cons) {
+		for (sl = 0; sl < 4; sl++) {
+			len += sprintf(page + len, PRINTK_HEADER
+				       " Sense(hex) %2d-%2d:",
+				       (8 * sl), ((8 * sl) + 7));
+
+			for (sct = 0; sct < 8; sct++) {
+				len += sprintf(page + len, " %02x",
+					       irb->ecw[8 * sl + sct]);
+			}
+			len += sprintf(page + len, "\n");
+		}
+	} else {
+		len += sprintf(page + len, PRINTK_HEADER
+			       " SORRY - NO VALID SENSE AVAILABLE\n");
+	}
+	printk(KERN_ERR "%s", page);
+
+	/* dump the Channel Program */
+	/* print first CCWs (maximum 8) */
+	act = req->cpaddr;
+        for (last = act; last->flags & (CCW_FLAG_CC | CCW_FLAG_DC); last++);
+	end = min(act + 8, last);
+	len = sprintf(page, PRINTK_HEADER " Related CP in req: %p\n", req);
+	while (act <= end) {
+		len += sprintf(page + len, PRINTK_HEADER
+			       " CCW %p: %08X %08X DAT:",
+			       act, ((int *) act)[0], ((int *) act)[1]);
+		for (count = 0; count < 32 && count < act->count;
+		     count += sizeof(int))
+			len += sprintf(page + len, " %08X",
+				       ((int *) (addr_t) act->cda)
+				       [(count>>2)]);
+		len += sprintf(page + len, "\n");
+		act++;
+	}
+	printk(KERN_ERR "%s", page);
+
+
+	/* print failing CCW area */
+	len = 0;
+	if (act <  ((struct ccw1 *)(addr_t) irb->scsw.cmd.cpa) - 2) {
+		act = ((struct ccw1 *)(addr_t) irb->scsw.cmd.cpa) - 2;
+		len += sprintf(page + len, PRINTK_HEADER "......\n");
+	}
+	end = min((struct ccw1 *)(addr_t) irb->scsw.cmd.cpa + 2, last);
+	while (act <= end) {
+		len += sprintf(page + len, PRINTK_HEADER
+			       " CCW %p: %08X %08X DAT:",
+			       act, ((int *) act)[0], ((int *) act)[1]);
+		for (count = 0; count < 32 && count < act->count;
+		     count += sizeof(int))
+			len += sprintf(page + len, " %08X",
+				       ((int *) (addr_t) act->cda)
+				       [(count>>2)]);
+		len += sprintf(page + len, "\n");
+		act++;
+	}
+
+	/* print last CCWs */
+	if (act <  last - 2) {
+		act = last - 2;
+		len += sprintf(page + len, PRINTK_HEADER "......\n");
+	}
+	while (act <= last) {
+		len += sprintf(page + len, PRINTK_HEADER
+			       " CCW %p: %08X %08X DAT:",
+			       act, ((int *) act)[0], ((int *) act)[1]);
+		for (count = 0; count < 32 && count < act->count;
+		     count += sizeof(int))
+			len += sprintf(page + len, " %08X",
+				       ((int *) (addr_t) act->cda)
+				       [(count>>2)]);
+		len += sprintf(page + len, "\n");
+		act++;
+	}
+	if (len > 0)
+		printk(KERN_ERR "%s", page);
+	free_page((unsigned long) page);
+}
+
+/*
+ * max_blocks is dependent on the amount of storage that is available
+ * in the static io buffer for each device. Currently each device has
+ * 8192 bytes (=2 pages). For 64 bit one dasd_mchunkt_t structure has
+ * 24 bytes, the struct dasd_ccw_req has 136 bytes and each block can use
+ * up to 16 bytes (8 for the ccw and 8 for the idal pointer). In
+ * addition we have one define extent ccw + 16 bytes of data and a
+ * locate record ccw for each block (stupid devices!) + 16 bytes of data.
+ * That makes:
+ * (8192 - 24 - 136 - 8 - 16) / 40 = 200.2 blocks at maximum.
+ * We want to fit two into the available memory so that we can immediately
+ * start the next request if one finishes off. That makes 100.1 blocks
+ * for one request. Give a little safety and the result is 96.
+ */
+static struct dasd_discipline dasd_fba_discipline = {
+	.owner = THIS_MODULE,
+	.name = "FBA ",
+	.ebcname = "FBA ",
+	.max_blocks = 96,
+	.check_device = dasd_fba_check_characteristics,
+	.do_analysis = dasd_fba_do_analysis,
+	.verify_path = dasd_generic_verify_path,
+	.fill_geometry = dasd_fba_fill_geometry,
+	.start_IO = dasd_start_IO,
+	.term_IO = dasd_term_IO,
+	.handle_terminated_request = dasd_fba_handle_terminated_request,
+	.erp_action = dasd_fba_erp_action,
+	.erp_postaction = dasd_fba_erp_postaction,
+	.check_for_device_change = dasd_fba_check_for_device_change,
+	.build_cp = dasd_fba_build_cp,
+	.free_cp = dasd_fba_free_cp,
+	.dump_sense = dasd_fba_dump_sense,
+	.dump_sense_dbf = dasd_fba_dump_sense_dbf,
+	.fill_info = dasd_fba_fill_info,
+};
+
+static int __init
+dasd_fba_init(void)
+{
+	int ret;
+
+	ASCEBC(dasd_fba_discipline.ebcname, 4);
+	ret = ccw_driver_register(&dasd_fba_driver);
+	if (!ret)
+		wait_for_device_probe();
+
+	return ret;
+}
+
+static void __exit
+dasd_fba_cleanup(void)
+{
+	ccw_driver_unregister(&dasd_fba_driver);
+}
+
+module_init(dasd_fba_init);
+module_exit(dasd_fba_cleanup);

drivers/s390/block/dasd_fba.h

@@ -0,0 +1,71 @@
+/*
+ * Author(s)......: Holger Smolinski <Holger.Smolinski@de.ibm.com>
+ * Bugreports.to..: <Linux390@de.ibm.com>
+ * Coypright IBM Corp. 1999, 2000
+ *
+ */
+
+#ifndef DASD_FBA_H
+#define DASD_FBA_H
+
+struct DE_fba_data {
+	struct {
+		unsigned char perm:2;	/* Permissions on this extent */
+		unsigned char zero:2;	/* Must be zero */
+		unsigned char da:1;	/* usually zero */
+		unsigned char diag:1;	/* allow diagnose */
+		unsigned char zero2:2;	/* zero */
+	} __attribute__ ((packed)) mask;
+	__u8 zero;		/* Must be zero */
+	__u16 blk_size;		/* Blocksize */
+	__u32 ext_loc;		/* Extent locator */
+	__u32 ext_beg;		/* logical number of block 0 in extent */
+	__u32 ext_end;		/* logocal number of last block in extent */
+} __attribute__ ((packed));
+
+struct LO_fba_data {
+	struct {
+		unsigned char zero:4;
+		unsigned char cmd:4;
+	} __attribute__ ((packed)) operation;
+	__u8 auxiliary;
+	__u16 blk_ct;
+	__u32 blk_nr;
+} __attribute__ ((packed));
+
+struct dasd_fba_characteristics {
+	union {
+		__u8 c;
+		struct {
+			unsigned char reserved:1;
+			unsigned char overrunnable:1;
+			unsigned char burst_byte:1;
+			unsigned char data_chain:1;
+			unsigned char zeros:4;
+		} __attribute__ ((packed)) bits;
+	} __attribute__ ((packed)) mode;
+	union {
+		__u8 c;
+		struct {
+			unsigned char zero0:1;
+			unsigned char removable:1;
+			unsigned char shared:1;
+			unsigned char zero1:1;
+			unsigned char mam:1;
+			unsigned char zeros:3;
+		} __attribute__ ((packed)) bits;
+	} __attribute__ ((packed)) features;
+	__u8 dev_class;
+	__u8 unit_type;
+	__u16 blk_size;
+	__u32 blk_per_cycl;
+	__u32 blk_per_bound;
+	__u32 blk_bdsa;
+	__u32 reserved0;
+	__u16 reserved1;
+	__u16 blk_ce;
+	__u32 reserved2;
+	__u16 reserved3;
+} __attribute__ ((packed));
+
+#endif				/* DASD_FBA_H */

drivers/s390/block/dasd_genhd.c

@@ -0,0 +1,178 @@
+/*
+ * Author(s)......: Holger Smolinski <Holger.Smolinski@de.ibm.com>
+ *		    Horst Hummel <Horst.Hummel@de.ibm.com>
+ *		    Carsten Otte <Cotte@de.ibm.com>
+ *		    Martin Schwidefsky <schwidefsky@de.ibm.com>
+ * Bugreports.to..: <Linux390@de.ibm.com>
+ * Copyright IBM Corp. 1999, 2001
+ *
+ * gendisk related functions for the dasd driver.
+ *
+ */
+
+#define KMSG_COMPONENT "dasd"
+
+#include <linux/interrupt.h>
+#include <linux/fs.h>
+#include <linux/blkpg.h>
+
+#include <asm/uaccess.h>
+
+/* This is ugly... */
+#define PRINTK_HEADER "dasd_gendisk:"
+
+#include "dasd_int.h"
+
+/*
+ * Allocate and register gendisk structure for device.
+ */
+int dasd_gendisk_alloc(struct dasd_block *block)
+{
+	struct gendisk *gdp;
+	struct dasd_device *base;
+	int len;
+
+	/* Make sure the minor for this device exists. */
+	base = block->base;
+	if (base->devindex >= DASD_PER_MAJOR)
+		return -EBUSY;
+
+	gdp = alloc_disk(1 << DASD_PARTN_BITS);
+	if (!gdp)
+		return -ENOMEM;
+
+	/* Initialize gendisk structure. */
+	gdp->major = DASD_MAJOR;
+	gdp->first_minor = base->devindex << DASD_PARTN_BITS;
+	gdp->fops = &dasd_device_operations;
+	gdp->driverfs_dev = &base->cdev->dev;
+
+	/*
+	 * Set device name.
+	 *   dasda - dasdz : 26 devices
+	 *   dasdaa - dasdzz : 676 devices, added up = 702
+	 *   dasdaaa - dasdzzz : 17576 devices, added up = 18278
+	 *   dasdaaaa - dasdzzzz : 456976 devices, added up = 475252
+	 */
+	len = sprintf(gdp->disk_name, "dasd");
+	if (base->devindex > 25) {
+		if (base->devindex > 701) {
+			if (base->devindex > 18277)
+			        len += sprintf(gdp->disk_name + len, "%c",
+					       'a'+(((base->devindex-18278)
+						     /17576)%26));
+			len += sprintf(gdp->disk_name + len, "%c",
+				       'a'+(((base->devindex-702)/676)%26));
+		}
+		len += sprintf(gdp->disk_name + len, "%c",
+			       'a'+(((base->devindex-26)/26)%26));
+	}
+	len += sprintf(gdp->disk_name + len, "%c", 'a'+(base->devindex%26));
+
+	if (base->features & DASD_FEATURE_READONLY ||
+	    test_bit(DASD_FLAG_DEVICE_RO, &base->flags))
+		set_disk_ro(gdp, 1);
+	dasd_add_link_to_gendisk(gdp, base);
+	gdp->queue = block->request_queue;
+	block->gdp = gdp;
+	set_capacity(block->gdp, 0);
+	add_disk(block->gdp);
+	return 0;
+}
+
+/*
+ * Unregister and free gendisk structure for device.
+ */
+void dasd_gendisk_free(struct dasd_block *block)
+{
+	if (block->gdp) {
+		del_gendisk(block->gdp);
+		block->gdp->private_data = NULL;
+		put_disk(block->gdp);
+		block->gdp = NULL;
+	}
+}
+
+/*
+ * Trigger a partition detection.
+ */
+int dasd_scan_partitions(struct dasd_block *block)
+{
+	struct block_device *bdev;
+
+	bdev = bdget_disk(block->gdp, 0);
+	if (!bdev || blkdev_get(bdev, FMODE_READ, NULL) < 0)
+		return -ENODEV;
+	/*
+	 * See fs/partition/check.c:register_disk,rescan_partitions
+	 * Can't call rescan_partitions directly. Use ioctl.
+	 */
+	ioctl_by_bdev(bdev, BLKRRPART, 0);
+	/*
+	 * Since the matching blkdev_put call to the blkdev_get in
+	 * this function is not called before dasd_destroy_partitions
+	 * the offline open_count limit needs to be increased from
+	 * 0 to 1. This is done by setting device->bdev (see
+	 * dasd_generic_set_offline). As long as the partition
+	 * detection is running no offline should be allowed. That
+	 * is why the assignment to device->bdev is done AFTER
+	 * the BLKRRPART ioctl.
+	 */
+	block->bdev = bdev;
+	return 0;
+}
+
+/*
+ * Remove all inodes in the system for a device, delete the
+ * partitions and make device unusable by setting its size to zero.
+ */
+void dasd_destroy_partitions(struct dasd_block *block)
+{
+	/* The two structs have 168/176 byte on 31/64 bit. */
+	struct blkpg_partition bpart;
+	struct blkpg_ioctl_arg barg;
+	struct block_device *bdev;
+
+	/*
+	 * Get the bdev pointer from the device structure and clear
+	 * device->bdev to lower the offline open_count limit again.
+	 */
+	bdev = block->bdev;
+	block->bdev = NULL;
+
+	/*
+	 * See fs/partition/check.c:delete_partition
+	 * Can't call delete_partitions directly. Use ioctl.
+	 * The ioctl also does locking and invalidation.
+	 */
+	memset(&bpart, 0, sizeof(struct blkpg_partition));
+	memset(&barg, 0, sizeof(struct blkpg_ioctl_arg));
+	barg.data = (void __force __user *) &bpart;
+	barg.op = BLKPG_DEL_PARTITION;
+	for (bpart.pno = block->gdp->minors - 1; bpart.pno > 0; bpart.pno--)
+		ioctl_by_bdev(bdev, BLKPG, (unsigned long) &barg);
+
+	invalidate_partition(block->gdp, 0);
+	/* Matching blkdev_put to the blkdev_get in dasd_scan_partitions. */
+	blkdev_put(bdev, FMODE_READ);
+	set_capacity(block->gdp, 0);
+}
+
+int dasd_gendisk_init(void)
+{
+	int rc;
+
+	/* Register to static dasd major 94 */
+	rc = register_blkdev(DASD_MAJOR, "dasd");
+	if (rc != 0) {
+		pr_warning("Registering the device driver with major number "
+			   "%d failed\n", DASD_MAJOR);
+		return rc;
+	}
+	return 0;
+}
+
+void dasd_gendisk_exit(void)
+{
+	unregister_blkdev(DASD_MAJOR, "dasd");
+}

drivers/s390/block/dasd_int.h

@@ -0,0 +1,815 @@
+/*
+ * Author(s)......: Holger Smolinski <Holger.Smolinski@de.ibm.com>
+ *		    Horst Hummel <Horst.Hummel@de.ibm.com>
+ *		    Martin Schwidefsky <schwidefsky@de.ibm.com>
+ * Bugreports.to..: <Linux390@de.ibm.com>
+ * Copyright IBM Corp. 1999, 2009
+ */
+
+#ifndef DASD_INT_H
+#define DASD_INT_H
+
+/* we keep old device allocation scheme; IOW, minors are still in 0..255 */
+#define DASD_PER_MAJOR (1U << (MINORBITS - DASD_PARTN_BITS))
+#define DASD_PARTN_MASK ((1 << DASD_PARTN_BITS) - 1)
+
+/*
+ * States a dasd device can have:
+ *   new: the dasd_device structure is allocated.
+ *   known: the discipline for the device is identified.
+ *   basic: the device can do basic i/o.
+ *   unfmt: the device could not be analyzed (format is unknown).
+ *   ready: partition detection is done and the device is can do block io.
+ *   online: the device accepts requests from the block device queue.
+ *
+ * Things to do for startup state transitions:
+ *   new -> known: find discipline for the device and create devfs entries.
+ *   known -> basic: request irq line for the device.
+ *   basic -> ready: do the initial analysis, e.g. format detection,
+ *                   do block device setup and detect partitions.
+ *   ready -> online: schedule the device tasklet.
+ * Things to do for shutdown state transitions:
+ *   online -> ready: just set the new device state.
+ *   ready -> basic: flush requests from the block device layer, clear
+ *                   partition information and reset format information.
+ *   basic -> known: terminate all requests and free irq.
+ *   known -> new: remove devfs entries and forget discipline.
+ */
+
+#define DASD_STATE_NEW	  0
+#define DASD_STATE_KNOWN  1
+#define DASD_STATE_BASIC  2
+#define DASD_STATE_UNFMT  3
+#define DASD_STATE_READY  4
+#define DASD_STATE_ONLINE 5
+
+#include <linux/module.h>
+#include <linux/wait.h>
+#include <linux/blkdev.h>
+#include <linux/genhd.h>
+#include <linux/hdreg.h>
+#include <linux/interrupt.h>
+#include <linux/log2.h>
+#include <asm/ccwdev.h>
+#include <linux/workqueue.h>
+#include <asm/debug.h>
+#include <asm/dasd.h>
+#include <asm/idals.h>
+
+/* DASD discipline magic */
+#define DASD_ECKD_MAGIC 0xC5C3D2C4
+#define DASD_DIAG_MAGIC 0xC4C9C1C7
+#define DASD_FBA_MAGIC 0xC6C2C140
+
+/*
+ * SECTION: Type definitions
+ */
+struct dasd_device;
+struct dasd_block;
+
+/* BIT DEFINITIONS FOR SENSE DATA */
+#define DASD_SENSE_BIT_0 0x80
+#define DASD_SENSE_BIT_1 0x40
+#define DASD_SENSE_BIT_2 0x20
+#define DASD_SENSE_BIT_3 0x10
+
+/* BIT DEFINITIONS FOR SIM SENSE */
+#define DASD_SIM_SENSE 0x0F
+#define DASD_SIM_MSG_TO_OP 0x03
+#define DASD_SIM_LOG 0x0C
+
+/* lock class for nested cdev lock */
+#define CDEV_NESTED_FIRST 1
+#define CDEV_NESTED_SECOND 2
+
+/*
+ * SECTION: MACROs for klogd and s390 debug feature (dbf)
+ */
+#define DBF_DEV_EVENT(d_level, d_device, d_str, d_data...) \
+do { \
+	debug_sprintf_event(d_device->debug_area, \
+			    d_level, \
+			    d_str "\n", \
+			    d_data); \
+} while(0)
+
+#define DBF_DEV_EXC(d_level, d_device, d_str, d_data...) \
+do { \
+	debug_sprintf_exception(d_device->debug_area, \
+				d_level, \
+				d_str "\n", \
+				d_data); \
+} while(0)
+
+#define DBF_EVENT(d_level, d_str, d_data...)\
+do { \
+	debug_sprintf_event(dasd_debug_area, \
+			    d_level,\
+			    d_str "\n", \
+			    d_data); \
+} while(0)
+
+#define DBF_EVENT_DEVID(d_level, d_cdev, d_str, d_data...)	\
+do { \
+	struct ccw_dev_id __dev_id;			\
+	ccw_device_get_id(d_cdev, &__dev_id);		\
+	debug_sprintf_event(dasd_debug_area,		\
+			    d_level,					\
+			    "0.%x.%04x " d_str "\n",			\
+			    __dev_id.ssid, __dev_id.devno, d_data);	\
+} while (0)
+
+#define DBF_EXC(d_level, d_str, d_data...)\
+do { \
+	debug_sprintf_exception(dasd_debug_area, \
+				d_level,\
+				d_str "\n", \
+				d_data); \
+} while(0)
+
+/* limit size for an errorstring */
+#define ERRORLENGTH 30
+
+/* definition of dbf debug levels */
+#define	DBF_EMERG	0	/* system is unusable			*/
+#define	DBF_ALERT	1	/* action must be taken immediately	*/
+#define	DBF_CRIT	2	/* critical conditions			*/
+#define	DBF_ERR		3	/* error conditions			*/
+#define	DBF_WARNING	4	/* warning conditions			*/
+#define	DBF_NOTICE	5	/* normal but significant condition	*/
+#define	DBF_INFO	6	/* informational			*/
+#define	DBF_DEBUG	6	/* debug-level messages			*/
+
+/* messages to be written via klogd and dbf */
+#define DEV_MESSAGE(d_loglevel,d_device,d_string,d_args...)\
+do { \
+	printk(d_loglevel PRINTK_HEADER " %s: " d_string "\n", \
+	       dev_name(&d_device->cdev->dev), d_args); \
+	DBF_DEV_EVENT(DBF_ALERT, d_device, d_string, d_args); \
+} while(0)
+
+#define MESSAGE(d_loglevel,d_string,d_args...)\
+do { \
+	printk(d_loglevel PRINTK_HEADER " " d_string "\n", d_args); \
+	DBF_EVENT(DBF_ALERT, d_string, d_args); \
+} while(0)
+
+/* messages to be written via klogd only */
+#define DEV_MESSAGE_LOG(d_loglevel,d_device,d_string,d_args...)\
+do { \
+	printk(d_loglevel PRINTK_HEADER " %s: " d_string "\n", \
+	       dev_name(&d_device->cdev->dev), d_args); \
+} while(0)
+
+#define MESSAGE_LOG(d_loglevel,d_string,d_args...)\
+do { \
+	printk(d_loglevel PRINTK_HEADER " " d_string "\n", d_args); \
+} while(0)
+
+struct dasd_ccw_req {
+	unsigned int magic;		/* Eye catcher */
+	struct list_head devlist;	/* for dasd_device request queue */
+	struct list_head blocklist;	/* for dasd_block request queue */
+
+	/* Where to execute what... */
+	struct dasd_block *block;	/* the originating block device */
+	struct dasd_device *memdev;	/* the device used to allocate this */
+	struct dasd_device *startdev;	/* device the request is started on */
+	struct dasd_device *basedev;	/* base device if no block->base */
+	void *cpaddr;			/* address of ccw or tcw */
+	unsigned char cpmode;		/* 0 = cmd mode, 1 = itcw */
+	char status;			/* status of this request */
+	short retries;			/* A retry counter */
+	unsigned long flags;        	/* flags of this request */
+
+	/* ... and how */
+	unsigned long starttime;	/* jiffies time of request start */
+	unsigned long expires;		/* expiration period in jiffies */
+	char lpm;			/* logical path mask */
+	void *data;			/* pointer to data area */
+
+	/* these are important for recovering erroneous requests          */
+	int intrc;			/* internal error, e.g. from start_IO */
+	struct irb irb;			/* device status in case of an error */
+	struct dasd_ccw_req *refers;	/* ERP-chain queueing. */
+	void *function; 		/* originating ERP action */
+
+	/* these are for statistics only */
+	unsigned long long buildclk;	/* TOD-clock of request generation */
+	unsigned long long startclk;	/* TOD-clock of request start */
+	unsigned long long stopclk;	/* TOD-clock of request interrupt */
+	unsigned long long endclk;	/* TOD-clock of request termination */
+
+        /* Callback that is called after reaching final status. */
+	void (*callback)(struct dasd_ccw_req *, void *data);
+	void *callback_data;
+};
+
+/*
+ * dasd_ccw_req -> status can be:
+ */
+#define DASD_CQR_FILLED 	0x00	/* request is ready to be processed */
+#define DASD_CQR_DONE		0x01	/* request is completed successfully */
+#define DASD_CQR_NEED_ERP	0x02	/* request needs recovery action */
+#define DASD_CQR_IN_ERP 	0x03	/* request is in recovery */
+#define DASD_CQR_FAILED 	0x04	/* request is finally failed */
+#define DASD_CQR_TERMINATED	0x05	/* request was stopped by driver */
+
+#define DASD_CQR_QUEUED 	0x80	/* request is queued to be processed */
+#define DASD_CQR_IN_IO		0x81	/* request is currently in IO */
+#define DASD_CQR_ERROR		0x82	/* request is completed with error */
+#define DASD_CQR_CLEAR_PENDING	0x83	/* request is clear pending */
+#define DASD_CQR_CLEARED	0x84	/* request was cleared */
+#define DASD_CQR_SUCCESS	0x85	/* request was successful */
+
+/* default expiration time*/
+#define DASD_EXPIRES	  300
+#define DASD_EXPIRES_MAX  40000000
+#define DASD_RETRIES	  256
+#define DASD_RETRIES_MAX  32768
+
+/* per dasd_ccw_req flags */
+#define DASD_CQR_FLAGS_USE_ERP   0	/* use ERP for this request */
+#define DASD_CQR_FLAGS_FAILFAST  1	/* FAILFAST */
+#define DASD_CQR_VERIFY_PATH	 2	/* path verification request */
+#define DASD_CQR_ALLOW_SLOCK	 3	/* Try this request even when lock was
+					 * stolen. Should not be combined with
+					 * DASD_CQR_FLAGS_USE_ERP
+					 */
+
+/* Signature for error recovery functions. */
+typedef struct dasd_ccw_req *(*dasd_erp_fn_t) (struct dasd_ccw_req *);
+
+/*
+ * Unique identifier for dasd device.
+ */
+#define UA_NOT_CONFIGURED  0x00
+#define UA_BASE_DEVICE	   0x01
+#define UA_BASE_PAV_ALIAS  0x02
+#define UA_HYPER_PAV_ALIAS 0x03
+
+struct dasd_uid {
+	__u8 type;
+	char vendor[4];
+	char serial[15];
+	__u16 ssid;
+	__u8 real_unit_addr;
+	__u8 base_unit_addr;
+	char vduit[33];
+};
+
+/*
+ * the struct dasd_discipline is
+ * sth like a table of virtual functions, if you think of dasd_eckd
+ * inheriting dasd...
+ * no, currently we are not planning to reimplement the driver in C++
+ */
+struct dasd_discipline {
+	struct module *owner;
+	char ebcname[8];	/* a name used for tagging and printks */
+	char name[8];		/* a name used for tagging and printks */
+	int max_blocks;		/* maximum number of blocks to be chained */
+
+	struct list_head list;	/* used for list of disciplines */
+
+	/*
+	 * Device recognition functions. check_device is used to verify
+	 * the sense data and the information returned by read device
+	 * characteristics. It returns 0 if the discipline can be used
+	 * for the device in question. uncheck_device is called during
+	 * device shutdown to deregister a device from its discipline.
+	 */
+	int (*check_device) (struct dasd_device *);
+	void (*uncheck_device) (struct dasd_device *);
+
+	/*
+	 * do_analysis is used in the step from device state "basic" to
+	 * state "accept". It returns 0 if the device can be made ready,
+	 * it returns -EMEDIUMTYPE if the device can't be made ready or
+	 * -EAGAIN if do_analysis started a ccw that needs to complete
+	 * before the analysis may be repeated.
+	 */
+	int (*do_analysis) (struct dasd_block *);
+
+	/*
+	 * This function is called, when new paths become available.
+	 * Disciplins may use this callback to do necessary setup work,
+	 * e.g. verify that new path is compatible with the current
+	 * configuration.
+	 */
+	int (*verify_path)(struct dasd_device *, __u8);
+
+	/*
+	 * Last things to do when a device is set online, and first things
+	 * when it is set offline.
+	 */
+	int (*basic_to_ready) (struct dasd_device *);
+	int (*online_to_ready) (struct dasd_device *);
+	int (*basic_to_known)(struct dasd_device *);
+
+	/* (struct dasd_device *);
+	 * Device operation functions. build_cp creates a ccw chain for
+	 * a block device request, start_io starts the request and
+	 * term_IO cancels it (e.g. in case of a timeout). format_device
+	 * returns a ccw chain to be used to format the device.
+	 * handle_terminated_request allows to examine a cqr and prepare
+	 * it for retry.
+	 */
+	struct dasd_ccw_req *(*build_cp) (struct dasd_device *,
+					  struct dasd_block *,
+					  struct request *);
+	int (*start_IO) (struct dasd_ccw_req *);
+	int (*term_IO) (struct dasd_ccw_req *);
+	void (*handle_terminated_request) (struct dasd_ccw_req *);
+	int (*format_device) (struct dasd_device *,
+			      struct format_data_t *, int enable_pav);
+	int (*free_cp) (struct dasd_ccw_req *, struct request *);
+
+	/*
+	 * Error recovery functions. examine_error() returns a value that
+	 * indicates what to do for an error condition. If examine_error()
+	 * returns 'dasd_era_recover' erp_action() is called to create a
+	 * special error recovery ccw. erp_postaction() is called after
+	 * an error recovery ccw has finished its execution. dump_sense
+	 * is called for every error condition to print the sense data
+	 * to the console.
+	 */
+	dasd_erp_fn_t(*erp_action) (struct dasd_ccw_req *);
+	dasd_erp_fn_t(*erp_postaction) (struct dasd_ccw_req *);
+	void (*dump_sense) (struct dasd_device *, struct dasd_ccw_req *,
+			    struct irb *);
+	void (*dump_sense_dbf) (struct dasd_device *, struct irb *, char *);
+	void (*check_for_device_change) (struct dasd_device *,
+					 struct dasd_ccw_req *,
+					 struct irb *);
+
+        /* i/o control functions. */
+	int (*fill_geometry) (struct dasd_block *, struct hd_geometry *);
+	int (*fill_info) (struct dasd_device *, struct dasd_information2_t *);
+	int (*ioctl) (struct dasd_block *, unsigned int, void __user *);
+
+	/* suspend/resume functions */
+	int (*freeze) (struct dasd_device *);
+	int (*restore) (struct dasd_device *);
+
+	/* reload device after state change */
+	int (*reload) (struct dasd_device *);
+
+	int (*get_uid) (struct dasd_device *, struct dasd_uid *);
+	void (*kick_validate) (struct dasd_device *);
+	int (*check_attention)(struct dasd_device *, __u8);
+};
+
+extern struct dasd_discipline *dasd_diag_discipline_pointer;
+
+/*
+ * Notification numbers for extended error reporting notifications:
+ * The DASD_EER_DISABLE notification is sent before a dasd_device (and it's
+ * eer pointer) is freed. The error reporting module needs to do all necessary
+ * cleanup steps.
+ * The DASD_EER_TRIGGER notification sends the actual error reports (triggers).
+ */
+#define DASD_EER_DISABLE 0
+#define DASD_EER_TRIGGER 1
+
+/* Trigger IDs for extended error reporting DASD_EER_TRIGGER notification */
+#define DASD_EER_FATALERROR  1
+#define DASD_EER_NOPATH      2
+#define DASD_EER_STATECHANGE 3
+#define DASD_EER_PPRCSUSPEND 4
+
+struct dasd_path {
+	__u8 opm;
+	__u8 tbvpm;
+	__u8 ppm;
+	__u8 npm;
+	/* paths that are not used because of a special condition */
+	__u8 cablepm; /* miss-cabled */
+	__u8 hpfpm;   /* the HPF requirements of the other paths are not met */
+	__u8 cuirpm;  /* CUIR varied offline */
+};
+
+struct dasd_profile_info {
+	/* legacy part of profile data, as in dasd_profile_info_t */
+	unsigned int dasd_io_reqs;	 /* number of requests processed */
+	unsigned int dasd_io_sects;	 /* number of sectors processed */
+	unsigned int dasd_io_secs[32];	 /* histogram of request's sizes */
+	unsigned int dasd_io_times[32];	 /* histogram of requests's times */
+	unsigned int dasd_io_timps[32];	 /* h. of requests's times per sector */
+	unsigned int dasd_io_time1[32];	 /* hist. of time from build to start */
+	unsigned int dasd_io_time2[32];	 /* hist. of time from start to irq */
+	unsigned int dasd_io_time2ps[32]; /* hist. of time from start to irq */
+	unsigned int dasd_io_time3[32];	 /* hist. of time from irq to end */
+	unsigned int dasd_io_nr_req[32]; /* hist. of # of requests in chanq */
+
+	/* new data */
+	struct timespec starttod;	   /* time of start or last reset */
+	unsigned int dasd_io_alias;	   /* requests using an alias */
+	unsigned int dasd_io_tpm;	   /* requests using transport mode */
+	unsigned int dasd_read_reqs;	   /* total number of read  requests */
+	unsigned int dasd_read_sects;	   /* total number read sectors */
+	unsigned int dasd_read_alias;	   /* read request using an alias */
+	unsigned int dasd_read_tpm;	   /* read requests in transport mode */
+	unsigned int dasd_read_secs[32];   /* histogram of request's sizes */
+	unsigned int dasd_read_times[32];  /* histogram of requests's times */
+	unsigned int dasd_read_time1[32];  /* hist. time from build to start */
+	unsigned int dasd_read_time2[32];  /* hist. of time from start to irq */
+	unsigned int dasd_read_time3[32];  /* hist. of time from irq to end */
+	unsigned int dasd_read_nr_req[32]; /* hist. of # of requests in chanq */
+};
+
+struct dasd_profile {
+	struct dentry *dentry;
+	struct dasd_profile_info *data;
+	spinlock_t lock;
+};
+
+struct dasd_device {
+	/* Block device stuff. */
+	struct dasd_block *block;
+
+        unsigned int devindex;
+	unsigned long flags;	   /* per device flags */
+	unsigned short features;   /* copy of devmap-features (read-only!) */
+
+	/* extended error reporting stuff (eer) */
+	struct dasd_ccw_req *eer_cqr;
+
+	/* Device discipline stuff. */
+	struct dasd_discipline *discipline;
+	struct dasd_discipline *base_discipline;
+	char *private;
+	struct dasd_path path_data;
+
+	/* Device state and target state. */
+	int state, target;
+	struct mutex state_mutex;
+	int stopped;		/* device (ccw_device_start) was stopped */
+
+	/* reference count. */
+        atomic_t ref_count;
+
+	/* ccw queue and memory for static ccw/erp buffers. */
+	struct list_head ccw_queue;
+	spinlock_t mem_lock;
+	void *ccw_mem;
+	void *erp_mem;
+	struct list_head ccw_chunks;
+	struct list_head erp_chunks;
+
+	atomic_t tasklet_scheduled;
+        struct tasklet_struct tasklet;
+	struct work_struct kick_work;
+	struct work_struct restore_device;
+	struct work_struct reload_device;
+	struct work_struct kick_validate;
+	struct timer_list timer;
+
+	debug_info_t *debug_area;
+
+	struct ccw_device *cdev;
+
+	/* hook for alias management */
+	struct list_head alias_list;
+
+	/* default expiration time in s */
+	unsigned long default_expires;
+	unsigned long default_retries;
+
+	unsigned long blk_timeout;
+
+	struct dentry *debugfs_dentry;
+	struct dasd_profile profile;
+};
+
+struct dasd_block {
+	/* Block device stuff. */
+	struct gendisk *gdp;
+	struct request_queue *request_queue;
+	spinlock_t request_queue_lock;
+	struct block_device *bdev;
+	atomic_t open_count;
+
+	unsigned long long blocks; /* size of volume in blocks */
+	unsigned int bp_block;	   /* bytes per block */
+	unsigned int s2b_shift;	   /* log2 (bp_block/512) */
+
+	struct dasd_device *base;
+	struct list_head ccw_queue;
+	spinlock_t queue_lock;
+
+	atomic_t tasklet_scheduled;
+	struct tasklet_struct tasklet;
+	struct timer_list timer;
+
+	struct dentry *debugfs_dentry;
+	struct dasd_profile profile;
+};
+
+struct dasd_attention_data {
+	struct dasd_device *device;
+	__u8 lpum;
+};
+
+/* reasons why device (ccw_device_start) was stopped */
+#define DASD_STOPPED_NOT_ACC 1         /* not accessible */
+#define DASD_STOPPED_QUIESCE 2         /* Quiesced */
+#define DASD_STOPPED_PENDING 4         /* long busy */
+#define DASD_STOPPED_DC_WAIT 8         /* disconnected, wait */
+#define DASD_STOPPED_SU      16        /* summary unit check handling */
+#define DASD_STOPPED_PM      32        /* pm state transition */
+#define DASD_UNRESUMED_PM    64        /* pm resume failed state */
+
+/* per device flags */
+#define DASD_FLAG_OFFLINE	3	/* device is in offline processing */
+#define DASD_FLAG_EER_SNSS	4	/* A SNSS is required */
+#define DASD_FLAG_EER_IN_USE	5	/* A SNSS request is running */
+#define DASD_FLAG_DEVICE_RO	6	/* The device itself is read-only. Don't
+					 * confuse this with the user specified
+					 * read-only feature.
+					 */
+#define DASD_FLAG_IS_RESERVED	7	/* The device is reserved */
+#define DASD_FLAG_LOCK_STOLEN	8	/* The device lock was stolen */
+#define DASD_FLAG_SUSPENDED	9	/* The device was suspended */
+#define DASD_FLAG_SAFE_OFFLINE	10	/* safe offline processing requested*/
+#define DASD_FLAG_SAFE_OFFLINE_RUNNING	11	/* safe offline running */
+#define DASD_FLAG_ABORTALL	12	/* Abort all noretry requests */
+
+#define DASD_SLEEPON_START_TAG	((void *) 1)
+#define DASD_SLEEPON_END_TAG	((void *) 2)
+
+void dasd_put_device_wake(struct dasd_device *);
+
+/*
+ * Reference count inliners
+ */
+static inline void
+dasd_get_device(struct dasd_device *device)
+{
+	atomic_inc(&device->ref_count);
+}
+
+static inline void
+dasd_put_device(struct dasd_device *device)
+{
+	if (atomic_dec_return(&device->ref_count) == 0)
+		dasd_put_device_wake(device);
+}
+
+/*
+ * The static memory in ccw_mem and erp_mem is managed by a sorted
+ * list of free memory chunks.
+ */
+struct dasd_mchunk
+{
+	struct list_head list;
+	unsigned long size;
+} __attribute__ ((aligned(8)));
+
+static inline void
+dasd_init_chunklist(struct list_head *chunk_list, void *mem,
+		    unsigned long size)
+{
+	struct dasd_mchunk *chunk;
+
+	INIT_LIST_HEAD(chunk_list);
+	chunk = (struct dasd_mchunk *) mem;
+	chunk->size = size - sizeof(struct dasd_mchunk);
+	list_add(&chunk->list, chunk_list);
+}
+
+static inline void *
+dasd_alloc_chunk(struct list_head *chunk_list, unsigned long size)
+{
+	struct dasd_mchunk *chunk, *tmp;
+
+	size = (size + 7L) & -8L;
+	list_for_each_entry(chunk, chunk_list, list) {
+		if (chunk->size < size)
+			continue;
+		if (chunk->size > size + sizeof(struct dasd_mchunk)) {
+			char *endaddr = (char *) (chunk + 1) + chunk->size;
+			tmp = (struct dasd_mchunk *) (endaddr - size) - 1;
+			tmp->size = size;
+			chunk->size -= size + sizeof(struct dasd_mchunk);
+			chunk = tmp;
+		} else
+			list_del(&chunk->list);
+		return (void *) (chunk + 1);
+	}
+	return NULL;
+}
+
+static inline void
+dasd_free_chunk(struct list_head *chunk_list, void *mem)
+{
+	struct dasd_mchunk *chunk, *tmp;
+	struct list_head *p, *left;
+
+	chunk = (struct dasd_mchunk *)
+		((char *) mem - sizeof(struct dasd_mchunk));
+	/* Find out the left neighbour in chunk_list. */
+	left = chunk_list;
+	list_for_each(p, chunk_list) {
+		if (list_entry(p, struct dasd_mchunk, list) > chunk)
+			break;
+		left = p;
+	}
+	/* Try to merge with right neighbour = next element from left. */
+	if (left->next != chunk_list) {
+		tmp = list_entry(left->next, struct dasd_mchunk, list);
+		if ((char *) (chunk + 1) + chunk->size == (char *) tmp) {
+			list_del(&tmp->list);
+			chunk->size += tmp->size + sizeof(struct dasd_mchunk);
+		}
+	}
+	/* Try to merge with left neighbour. */
+	if (left != chunk_list) {
+		tmp = list_entry(left, struct dasd_mchunk, list);
+		if ((char *) (tmp + 1) + tmp->size == (char *) chunk) {
+			tmp->size += chunk->size + sizeof(struct dasd_mchunk);
+			return;
+		}
+	}
+	__list_add(&chunk->list, left, left->next);
+}
+
+/*
+ * Check if bsize is in { 512, 1024, 2048, 4096 }
+ */
+static inline int
+dasd_check_blocksize(int bsize)
+{
+	if (bsize < 512 || bsize > 4096 || !is_power_of_2(bsize))
+		return -EMEDIUMTYPE;
+	return 0;
+}
+
+/* externals in dasd.c */
+#define DASD_PROFILE_OFF	 0
+#define DASD_PROFILE_ON 	 1
+#define DASD_PROFILE_GLOBAL_ONLY 2
+
+extern debug_info_t *dasd_debug_area;
+extern struct dasd_profile_info dasd_global_profile_data;
+extern unsigned int dasd_global_profile_level;
+extern const struct block_device_operations dasd_device_operations;
+
+extern struct kmem_cache *dasd_page_cache;
+
+struct dasd_ccw_req *
+dasd_kmalloc_request(int , int, int, struct dasd_device *);
+struct dasd_ccw_req *
+dasd_smalloc_request(int , int, int, struct dasd_device *);
+void dasd_kfree_request(struct dasd_ccw_req *, struct dasd_device *);
+void dasd_sfree_request(struct dasd_ccw_req *, struct dasd_device *);
+void dasd_wakeup_cb(struct dasd_ccw_req *, void *);
+
+static inline int
+dasd_kmalloc_set_cda(struct ccw1 *ccw, void *cda, struct dasd_device *device)
+{
+	return set_normalized_cda(ccw, cda);
+}
+
+struct dasd_device *dasd_alloc_device(void);
+void dasd_free_device(struct dasd_device *);
+
+struct dasd_block *dasd_alloc_block(void);
+void dasd_free_block(struct dasd_block *);
+
+enum blk_eh_timer_return dasd_times_out(struct request *req);
+
+void dasd_enable_device(struct dasd_device *);
+void dasd_set_target_state(struct dasd_device *, int);
+void dasd_kick_device(struct dasd_device *);
+void dasd_restore_device(struct dasd_device *);
+void dasd_reload_device(struct dasd_device *);
+
+void dasd_add_request_head(struct dasd_ccw_req *);
+void dasd_add_request_tail(struct dasd_ccw_req *);
+int  dasd_start_IO(struct dasd_ccw_req *);
+int  dasd_term_IO(struct dasd_ccw_req *);
+void dasd_schedule_device_bh(struct dasd_device *);
+void dasd_schedule_block_bh(struct dasd_block *);
+int  dasd_sleep_on(struct dasd_ccw_req *);
+int  dasd_sleep_on_queue(struct list_head *);
+int  dasd_sleep_on_immediatly(struct dasd_ccw_req *);
+int  dasd_sleep_on_interruptible(struct dasd_ccw_req *);
+void dasd_device_set_timer(struct dasd_device *, int);
+void dasd_device_clear_timer(struct dasd_device *);
+void dasd_block_set_timer(struct dasd_block *, int);
+void dasd_block_clear_timer(struct dasd_block *);
+int  dasd_cancel_req(struct dasd_ccw_req *);
+int dasd_flush_device_queue(struct dasd_device *);
+int dasd_generic_probe (struct ccw_device *, struct dasd_discipline *);
+void dasd_generic_remove (struct ccw_device *cdev);
+int dasd_generic_set_online(struct ccw_device *, struct dasd_discipline *);
+int dasd_generic_set_offline (struct ccw_device *cdev);
+int dasd_generic_notify(struct ccw_device *, int);
+int dasd_generic_last_path_gone(struct dasd_device *);
+int dasd_generic_path_operational(struct dasd_device *);
+void dasd_generic_shutdown(struct ccw_device *);
+
+void dasd_generic_handle_state_change(struct dasd_device *);
+int dasd_generic_pm_freeze(struct ccw_device *);
+int dasd_generic_restore_device(struct ccw_device *);
+enum uc_todo dasd_generic_uc_handler(struct ccw_device *, struct irb *);
+void dasd_generic_path_event(struct ccw_device *, int *);
+int dasd_generic_verify_path(struct dasd_device *, __u8);
+
+int dasd_generic_read_dev_chars(struct dasd_device *, int, void *, int);
+char *dasd_get_sense(struct irb *);
+
+void dasd_device_set_stop_bits(struct dasd_device *, int);
+void dasd_device_remove_stop_bits(struct dasd_device *, int);
+
+int dasd_device_is_ro(struct dasd_device *);
+
+void dasd_profile_reset(struct dasd_profile *);
+int dasd_profile_on(struct dasd_profile *);
+void dasd_profile_off(struct dasd_profile *);
+void dasd_global_profile_reset(void);
+char *dasd_get_user_string(const char __user *, size_t);
+
+/* externals in dasd_devmap.c */
+extern int dasd_max_devindex;
+extern int dasd_probeonly;
+extern int dasd_autodetect;
+extern int dasd_nopav;
+extern int dasd_nofcx;
+
+int dasd_devmap_init(void);
+void dasd_devmap_exit(void);
+
+struct dasd_device *dasd_create_device(struct ccw_device *);
+void dasd_delete_device(struct dasd_device *);
+
+int dasd_get_feature(struct ccw_device *, int);
+int dasd_set_feature(struct ccw_device *, int, int);
+
+int dasd_add_sysfs_files(struct ccw_device *);
+void dasd_remove_sysfs_files(struct ccw_device *);
+
+struct dasd_device *dasd_device_from_cdev(struct ccw_device *);
+struct dasd_device *dasd_device_from_cdev_locked(struct ccw_device *);
+struct dasd_device *dasd_device_from_devindex(int);
+
+void dasd_add_link_to_gendisk(struct gendisk *, struct dasd_device *);
+struct dasd_device *dasd_device_from_gendisk(struct gendisk *);
+
+int dasd_parse(void);
+int dasd_busid_known(const char *);
+
+/* externals in dasd_gendisk.c */
+int  dasd_gendisk_init(void);
+void dasd_gendisk_exit(void);
+int dasd_gendisk_alloc(struct dasd_block *);
+void dasd_gendisk_free(struct dasd_block *);
+int dasd_scan_partitions(struct dasd_block *);
+void dasd_destroy_partitions(struct dasd_block *);
+
+/* externals in dasd_ioctl.c */
+int  dasd_ioctl(struct block_device *, fmode_t, unsigned int, unsigned long);
+
+/* externals in dasd_proc.c */
+int dasd_proc_init(void);
+void dasd_proc_exit(void);
+
+/* externals in dasd_erp.c */
+struct dasd_ccw_req *dasd_default_erp_action(struct dasd_ccw_req *);
+struct dasd_ccw_req *dasd_default_erp_postaction(struct dasd_ccw_req *);
+struct dasd_ccw_req *dasd_alloc_erp_request(char *, int, int,
+					    struct dasd_device *);
+void dasd_free_erp_request(struct dasd_ccw_req *, struct dasd_device *);
+void dasd_log_sense(struct dasd_ccw_req *, struct irb *);
+void dasd_log_sense_dbf(struct dasd_ccw_req *cqr, struct irb *irb);
+
+/* externals in dasd_3990_erp.c */
+struct dasd_ccw_req *dasd_3990_erp_action(struct dasd_ccw_req *);
+void dasd_3990_erp_handle_sim(struct dasd_device *, char *);
+
+/* externals in dasd_eer.c */
+#ifdef CONFIG_DASD_EER
+int dasd_eer_init(void);
+void dasd_eer_exit(void);
+int dasd_eer_enable(struct dasd_device *);
+void dasd_eer_disable(struct dasd_device *);
+void dasd_eer_write(struct dasd_device *, struct dasd_ccw_req *cqr,
+		    unsigned int id);
+void dasd_eer_snss(struct dasd_device *);
+
+static inline int dasd_eer_enabled(struct dasd_device *device)
+{
+	return device->eer_cqr != NULL;
+}
+#else
+#define dasd_eer_init()		(0)
+#define dasd_eer_exit()		do { } while (0)
+#define dasd_eer_enable(d)	(0)
+#define dasd_eer_disable(d)	do { } while (0)
+#define dasd_eer_write(d,c,i)	do { } while (0)
+#define dasd_eer_snss(d)	do { } while (0)
+#define dasd_eer_enabled(d)	(0)
+#endif	/* CONFIG_DASD_ERR */
+
+#endif				/* DASD_H */

drivers/s390/block/dasd_ioctl.c

@@ -0,0 +1,579 @@
+/*
+ * Author(s)......: Holger Smolinski <Holger.Smolinski@de.ibm.com>
+ *		    Horst Hummel <Horst.Hummel@de.ibm.com>
+ *		    Carsten Otte <Cotte@de.ibm.com>
+ *		    Martin Schwidefsky <schwidefsky@de.ibm.com>
+ * Bugreports.to..: <Linux390@de.ibm.com>
+ * Copyright IBM Corp. 1999, 2001
+ *
+ * i/o controls for the dasd driver.
+ */
+
+#define KMSG_COMPONENT "dasd"
+
+#include <linux/interrupt.h>
+#include <linux/compat.h>
+#include <linux/major.h>
+#include <linux/fs.h>
+#include <linux/blkpg.h>
+#include <linux/slab.h>
+#include <asm/compat.h>
+#include <asm/ccwdev.h>
+#include <asm/schid.h>
+#include <asm/cmb.h>
+#include <asm/uaccess.h>
+
+/* This is ugly... */
+#define PRINTK_HEADER "dasd_ioctl:"
+
+#include "dasd_int.h"
+
+
+static int
+dasd_ioctl_api_version(void __user *argp)
+{
+	int ver = DASD_API_VERSION;
+	return put_user(ver, (int __user *)argp);
+}
+
+/*
+ * Enable device.
+ * used by dasdfmt after BIODASDDISABLE to retrigger blocksize detection
+ */
+static int
+dasd_ioctl_enable(struct block_device *bdev)
+{
+	struct dasd_device *base;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EACCES;
+
+	base = dasd_device_from_gendisk(bdev->bd_disk);
+	if (!base)
+		return -ENODEV;
+
+	dasd_enable_device(base);
+	/* Formatting the dasd device can change the capacity. */
+	mutex_lock(&bdev->bd_mutex);
+	i_size_write(bdev->bd_inode,
+		     (loff_t)get_capacity(base->block->gdp) << 9);
+	mutex_unlock(&bdev->bd_mutex);
+	dasd_put_device(base);
+	return 0;
+}
+
+/*
+ * Disable device.
+ * Used by dasdfmt. Disable I/O operations but allow ioctls.
+ */
+static int
+dasd_ioctl_disable(struct block_device *bdev)
+{
+	struct dasd_device *base;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EACCES;
+
+	base = dasd_device_from_gendisk(bdev->bd_disk);
+	if (!base)
+		return -ENODEV;
+	/*
+	 * Man this is sick. We don't do a real disable but only downgrade
+	 * the device to DASD_STATE_BASIC. The reason is that dasdfmt uses
+	 * BIODASDDISABLE to disable accesses to the device via the block
+	 * device layer but it still wants to do i/o on the device by
+	 * using the BIODASDFMT ioctl. Therefore the correct state for the
+	 * device is DASD_STATE_BASIC that allows to do basic i/o.
+	 */
+	dasd_set_target_state(base, DASD_STATE_BASIC);
+	/*
+	 * Set i_size to zero, since read, write, etc. check against this
+	 * value.
+	 */
+	mutex_lock(&bdev->bd_mutex);
+	i_size_write(bdev->bd_inode, 0);
+	mutex_unlock(&bdev->bd_mutex);
+	dasd_put_device(base);
+	return 0;
+}
+
+/*
+ * Quiesce device.
+ */
+static int dasd_ioctl_quiesce(struct dasd_block *block)
+{
+	unsigned long flags;
+	struct dasd_device *base;
+
+	base = block->base;
+	if (!capable (CAP_SYS_ADMIN))
+		return -EACCES;
+
+	pr_info("%s: The DASD has been put in the quiesce "
+		"state\n", dev_name(&base->cdev->dev));
+	spin_lock_irqsave(get_ccwdev_lock(base->cdev), flags);
+	dasd_device_set_stop_bits(base, DASD_STOPPED_QUIESCE);
+	spin_unlock_irqrestore(get_ccwdev_lock(base->cdev), flags);
+	return 0;
+}
+
+
+/*
+ * Resume device.
+ */
+static int dasd_ioctl_resume(struct dasd_block *block)
+{
+	unsigned long flags;
+	struct dasd_device *base;
+
+	base = block->base;
+	if (!capable (CAP_SYS_ADMIN))
+		return -EACCES;
+
+	pr_info("%s: I/O operations have been resumed "
+		"on the DASD\n", dev_name(&base->cdev->dev));
+	spin_lock_irqsave(get_ccwdev_lock(base->cdev), flags);
+	dasd_device_remove_stop_bits(base, DASD_STOPPED_QUIESCE);
+	spin_unlock_irqrestore(get_ccwdev_lock(base->cdev), flags);
+
+	dasd_schedule_block_bh(block);
+	return 0;
+}
+
+/*
+ * Abort all failfast I/O on a device.
+ */
+static int dasd_ioctl_abortio(struct dasd_block *block)
+{
+	unsigned long flags;
+	struct dasd_device *base;
+	struct dasd_ccw_req *cqr, *n;
+
+	base = block->base;
+	if (!capable(CAP_SYS_ADMIN))
+		return -EACCES;
+
+	if (test_and_set_bit(DASD_FLAG_ABORTALL, &base->flags))
+		return 0;
+	DBF_DEV_EVENT(DBF_NOTICE, base, "%s", "abortall flag set");
+
+	spin_lock_irqsave(&block->request_queue_lock, flags);
+	spin_lock(&block->queue_lock);
+	list_for_each_entry_safe(cqr, n, &block->ccw_queue, blocklist) {
+		if (test_bit(DASD_CQR_FLAGS_FAILFAST, &cqr->flags) &&
+		    cqr->callback_data &&
+		    cqr->callback_data != DASD_SLEEPON_START_TAG &&
+		    cqr->callback_data != DASD_SLEEPON_END_TAG) {
+			spin_unlock(&block->queue_lock);
+			blk_abort_request(cqr->callback_data);
+			spin_lock(&block->queue_lock);
+		}
+	}
+	spin_unlock(&block->queue_lock);
+	spin_unlock_irqrestore(&block->request_queue_lock, flags);
+
+	dasd_schedule_block_bh(block);
+	return 0;
+}
+
+/*
+ * Allow I/O on a device
+ */
+static int dasd_ioctl_allowio(struct dasd_block *block)
+{
+	struct dasd_device *base;
+
+	base = block->base;
+	if (!capable(CAP_SYS_ADMIN))
+		return -EACCES;
+
+	if (test_and_clear_bit(DASD_FLAG_ABORTALL, &base->flags))
+		DBF_DEV_EVENT(DBF_NOTICE, base, "%s", "abortall flag unset");
+
+	return 0;
+}
+
+/*
+ * performs formatting of _device_ according to _fdata_
+ * Note: The discipline's format_function is assumed to deliver formatting
+ * commands to format multiple units of the device. In terms of the ECKD
+ * devices this means CCWs are generated to format multiple tracks.
+ */
+static int
+dasd_format(struct dasd_block *block, struct format_data_t *fdata)
+{
+	struct dasd_device *base;
+	int enable_pav = 1;
+	int rc, retries;
+	int start, stop;
+
+	base = block->base;
+	if (base->discipline->format_device == NULL)
+		return -EPERM;
+
+	if (base->state != DASD_STATE_BASIC) {
+		pr_warn("%s: The DASD cannot be formatted while it is enabled\n",
+			dev_name(&base->cdev->dev));
+		return -EBUSY;
+	}
+
+	DBF_DEV_EVENT(DBF_NOTICE, base,
+		      "formatting units %u to %u (%u B blocks) flags %u",
+		      fdata->start_unit,
+		      fdata->stop_unit, fdata->blksize, fdata->intensity);
+
+	/* Since dasdfmt keeps the device open after it was disabled,
+	 * there still exists an inode for this device.
+	 * We must update i_blkbits, otherwise we might get errors when
+	 * enabling the device later.
+	 */
+	if (fdata->start_unit == 0) {
+		struct block_device *bdev = bdget_disk(block->gdp, 0);
+		bdev->bd_inode->i_blkbits = blksize_bits(fdata->blksize);
+		bdput(bdev);
+	}
+
+	retries = 255;
+	/* backup start- and endtrack for retries */
+	start = fdata->start_unit;
+	stop = fdata->stop_unit;
+	do {
+		rc = base->discipline->format_device(base, fdata, enable_pav);
+		if (rc) {
+			if (rc == -EAGAIN) {
+				retries--;
+				/* disable PAV in case of errors */
+				enable_pav = 0;
+				fdata->start_unit = start;
+				fdata->stop_unit = stop;
+			} else
+				return rc;
+		} else
+			/* success */
+			break;
+	} while (retries);
+
+	if (!retries)
+		return -EIO;
+	else
+		return 0;
+}
+
+/*
+ * Format device.
+ */
+static int
+dasd_ioctl_format(struct block_device *bdev, void __user *argp)
+{
+	struct dasd_device *base;
+	struct format_data_t fdata;
+	int rc;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EACCES;
+	if (!argp)
+		return -EINVAL;
+	base = dasd_device_from_gendisk(bdev->bd_disk);
+	if (!base)
+		return -ENODEV;
+	if (base->features & DASD_FEATURE_READONLY ||
+	    test_bit(DASD_FLAG_DEVICE_RO, &base->flags)) {
+		dasd_put_device(base);
+		return -EROFS;
+	}
+	if (copy_from_user(&fdata, argp, sizeof(struct format_data_t))) {
+		dasd_put_device(base);
+		return -EFAULT;
+	}
+	if (bdev != bdev->bd_contains) {
+		pr_warning("%s: The specified DASD is a partition and cannot "
+			   "be formatted\n",
+			   dev_name(&base->cdev->dev));
+		dasd_put_device(base);
+		return -EINVAL;
+	}
+	rc = dasd_format(base->block, &fdata);
+	dasd_put_device(base);
+	return rc;
+}
+
+#ifdef CONFIG_DASD_PROFILE
+/*
+ * Reset device profile information
+ */
+static int dasd_ioctl_reset_profile(struct dasd_block *block)
+{
+	dasd_profile_reset(&block->profile);
+	return 0;
+}
+
+/*
+ * Return device profile information
+ */
+static int dasd_ioctl_read_profile(struct dasd_block *block, void __user *argp)
+{
+	struct dasd_profile_info_t *data;
+	int rc = 0;
+
+	data = kmalloc(sizeof(*data), GFP_KERNEL);
+	if (!data)
+		return -ENOMEM;
+
+	spin_lock_bh(&block->profile.lock);
+	if (block->profile.data) {
+		data->dasd_io_reqs = block->profile.data->dasd_io_reqs;
+		data->dasd_io_sects = block->profile.data->dasd_io_sects;
+		memcpy(data->dasd_io_secs, block->profile.data->dasd_io_secs,
+		       sizeof(data->dasd_io_secs));
+		memcpy(data->dasd_io_times, block->profile.data->dasd_io_times,
+		       sizeof(data->dasd_io_times));
+		memcpy(data->dasd_io_timps, block->profile.data->dasd_io_timps,
+		       sizeof(data->dasd_io_timps));
+		memcpy(data->dasd_io_time1, block->profile.data->dasd_io_time1,
+		       sizeof(data->dasd_io_time1));
+		memcpy(data->dasd_io_time2, block->profile.data->dasd_io_time2,
+		       sizeof(data->dasd_io_time2));
+		memcpy(data->dasd_io_time2ps,
+		       block->profile.data->dasd_io_time2ps,
+		       sizeof(data->dasd_io_time2ps));
+		memcpy(data->dasd_io_time3, block->profile.data->dasd_io_time3,
+		       sizeof(data->dasd_io_time3));
+		memcpy(data->dasd_io_nr_req,
+		       block->profile.data->dasd_io_nr_req,
+		       sizeof(data->dasd_io_nr_req));
+		spin_unlock_bh(&block->profile.lock);
+	} else {
+		spin_unlock_bh(&block->profile.lock);
+		rc = -EIO;
+		goto out;
+	}
+	if (copy_to_user(argp, data, sizeof(*data)))
+		rc = -EFAULT;
+out:
+	kfree(data);
+	return rc;
+}
+#else
+static int dasd_ioctl_reset_profile(struct dasd_block *block)
+{
+	return -ENOTTY;
+}
+
+static int dasd_ioctl_read_profile(struct dasd_block *block, void __user *argp)
+{
+	return -ENOTTY;
+}
+#endif
+
+/*
+ * Return dasd information. Used for BIODASDINFO and BIODASDINFO2.
+ */
+static int dasd_ioctl_information(struct dasd_block *block,
+				  unsigned int cmd, void __user *argp)
+{
+	struct dasd_information2_t *dasd_info;
+	struct subchannel_id sch_id;
+	struct ccw_dev_id dev_id;
+	struct dasd_device *base;
+	struct ccw_device *cdev;
+	unsigned long flags;
+	int rc;
+
+	base = block->base;
+	if (!base->discipline || !base->discipline->fill_info)
+		return -EINVAL;
+
+	dasd_info = kzalloc(sizeof(struct dasd_information2_t), GFP_KERNEL);
+	if (dasd_info == NULL)
+		return -ENOMEM;
+
+	rc = base->discipline->fill_info(base, dasd_info);
+	if (rc) {
+		kfree(dasd_info);
+		return rc;
+	}
+
+	cdev = base->cdev;
+	ccw_device_get_id(cdev, &dev_id);
+	ccw_device_get_schid(cdev, &sch_id);
+
+	dasd_info->devno = dev_id.devno;
+	dasd_info->schid = sch_id.sch_no;
+	dasd_info->cu_type = cdev->id.cu_type;
+	dasd_info->cu_model = cdev->id.cu_model;
+	dasd_info->dev_type = cdev->id.dev_type;
+	dasd_info->dev_model = cdev->id.dev_model;
+	dasd_info->status = base->state;
+	/*
+	 * The open_count is increased for every opener, that includes
+	 * the blkdev_get in dasd_scan_partitions.
+	 * This must be hidden from user-space.
+	 */
+	dasd_info->open_count = atomic_read(&block->open_count);
+	if (!block->bdev)
+		dasd_info->open_count++;
+
+	/*
+	 * check if device is really formatted
+	 * LDL / CDL was returned by 'fill_info'
+	 */
+	if ((base->state < DASD_STATE_READY) ||
+	    (dasd_check_blocksize(block->bp_block)))
+		dasd_info->format = DASD_FORMAT_NONE;
+
+	dasd_info->features |=
+		((base->features & DASD_FEATURE_READONLY) != 0);
+
+	memcpy(dasd_info->type, base->discipline->name, 4);
+
+	if (block->request_queue->request_fn) {
+		struct list_head *l;
+#ifdef DASD_EXTENDED_PROFILING
+		{
+			struct list_head *l;
+			spin_lock_irqsave(&block->lock, flags);
+			list_for_each(l, &block->request_queue->queue_head)
+				dasd_info->req_queue_len++;
+			spin_unlock_irqrestore(&block->lock, flags);
+		}
+#endif				/* DASD_EXTENDED_PROFILING */
+		spin_lock_irqsave(get_ccwdev_lock(base->cdev), flags);
+		list_for_each(l, &base->ccw_queue)
+			dasd_info->chanq_len++;
+		spin_unlock_irqrestore(get_ccwdev_lock(base->cdev),
+				       flags);
+	}
+
+	rc = 0;
+	if (copy_to_user(argp, dasd_info,
+			 ((cmd == (unsigned int) BIODASDINFO2) ?
+			  sizeof(struct dasd_information2_t) :
+			  sizeof(struct dasd_information_t))))
+		rc = -EFAULT;
+	kfree(dasd_info);
+	return rc;
+}
+
+/*
+ * Set read only
+ */
+static int
+dasd_ioctl_set_ro(struct block_device *bdev, void __user *argp)
+{
+	struct dasd_device *base;
+	int intval, rc;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EACCES;
+	if (bdev != bdev->bd_contains)
+		// ro setting is not allowed for partitions
+		return -EINVAL;
+	if (get_user(intval, (int __user *)argp))
+		return -EFAULT;
+	base = dasd_device_from_gendisk(bdev->bd_disk);
+	if (!base)
+		return -ENODEV;
+	if (!intval && test_bit(DASD_FLAG_DEVICE_RO, &base->flags)) {
+		dasd_put_device(base);
+		return -EROFS;
+	}
+	set_disk_ro(bdev->bd_disk, intval);
+	rc = dasd_set_feature(base->cdev, DASD_FEATURE_READONLY, intval);
+	dasd_put_device(base);
+	return rc;
+}
+
+static int dasd_ioctl_readall_cmb(struct dasd_block *block, unsigned int cmd,
+				  struct cmbdata __user *argp)
+{
+	size_t size = _IOC_SIZE(cmd);
+	struct cmbdata data;
+	int ret;
+
+	ret = cmf_readall(block->base->cdev, &data);
+	if (!ret && copy_to_user(argp, &data, min(size, sizeof(*argp))))
+		return -EFAULT;
+	return ret;
+}
+
+int dasd_ioctl(struct block_device *bdev, fmode_t mode,
+	       unsigned int cmd, unsigned long arg)
+{
+	struct dasd_block *block;
+	struct dasd_device *base;
+	void __user *argp;
+	int rc;
+
+	if (is_compat_task())
+		argp = compat_ptr(arg);
+	else
+		argp = (void __user *)arg;
+
+	if ((_IOC_DIR(cmd) != _IOC_NONE) && !arg) {
+		PRINT_DEBUG("empty data ptr");
+		return -EINVAL;
+	}
+
+	base = dasd_device_from_gendisk(bdev->bd_disk);
+	if (!base)
+		return -ENODEV;
+	block = base->block;
+	rc = 0;
+	switch (cmd) {
+	case BIODASDDISABLE:
+		rc = dasd_ioctl_disable(bdev);
+		break;
+	case BIODASDENABLE:
+		rc = dasd_ioctl_enable(bdev);
+		break;
+	case BIODASDQUIESCE:
+		rc = dasd_ioctl_quiesce(block);
+		break;
+	case BIODASDRESUME:
+		rc = dasd_ioctl_resume(block);
+		break;
+	case BIODASDABORTIO:
+		rc = dasd_ioctl_abortio(block);
+		break;
+	case BIODASDALLOWIO:
+		rc = dasd_ioctl_allowio(block);
+		break;
+	case BIODASDFMT:
+		rc = dasd_ioctl_format(bdev, argp);
+		break;
+	case BIODASDINFO:
+		rc = dasd_ioctl_information(block, cmd, argp);
+		break;
+	case BIODASDINFO2:
+		rc = dasd_ioctl_information(block, cmd, argp);
+		break;
+	case BIODASDPRRD:
+		rc = dasd_ioctl_read_profile(block, argp);
+		break;
+	case BIODASDPRRST:
+		rc = dasd_ioctl_reset_profile(block);
+		break;
+	case BLKROSET:
+		rc = dasd_ioctl_set_ro(bdev, argp);
+		break;
+	case DASDAPIVER:
+		rc = dasd_ioctl_api_version(argp);
+		break;
+	case BIODASDCMFENABLE:
+		rc = enable_cmf(base->cdev);
+		break;
+	case BIODASDCMFDISABLE:
+		rc = disable_cmf(base->cdev);
+		break;
+	case BIODASDREADALLCMB:
+		rc = dasd_ioctl_readall_cmb(block, cmd, argp);
+		break;
+	default:
+		/* if the discipline has an ioctl method try it. */
+		rc = -ENOTTY;
+		if (base->discipline->ioctl)
+			rc = base->discipline->ioctl(block, cmd, argp);
+	}
+	dasd_put_device(base);
+	return rc;
+}

drivers/s390/block/dasd_proc.c

@@ -0,0 +1,376 @@
+/*
+ * Author(s)......: Holger Smolinski <Holger.Smolinski@de.ibm.com>
+ *		    Horst Hummel <Horst.Hummel@de.ibm.com>
+ *		    Carsten Otte <Cotte@de.ibm.com>
+ *		    Martin Schwidefsky <schwidefsky@de.ibm.com>
+ * Bugreports.to..: <Linux390@de.ibm.com>
+ * Coypright IBM Corp. 1999, 2002
+ *
+ * /proc interface for the dasd driver.
+ *
+ */
+
+#define KMSG_COMPONENT "dasd"
+
+#include <linux/ctype.h>
+#include <linux/slab.h>
+#include <linux/string.h>
+#include <linux/seq_file.h>
+#include <linux/vmalloc.h>
+#include <linux/proc_fs.h>
+
+#include <asm/debug.h>
+#include <asm/uaccess.h>
+
+/* This is ugly... */
+#define PRINTK_HEADER "dasd_proc:"
+
+#include "dasd_int.h"
+
+static struct proc_dir_entry *dasd_proc_root_entry = NULL;
+static struct proc_dir_entry *dasd_devices_entry = NULL;
+static struct proc_dir_entry *dasd_statistics_entry = NULL;
+
+static int
+dasd_devices_show(struct seq_file *m, void *v)
+{
+	struct dasd_device *device;
+	struct dasd_block *block;
+	char *substr;
+
+	device = dasd_device_from_devindex((unsigned long) v - 1);
+	if (IS_ERR(device))
+		return 0;
+	if (device->block)
+		block = device->block;
+	else {
+		dasd_put_device(device);
+		return 0;
+	}
+	/* Print device number. */
+	seq_printf(m, "%s", dev_name(&device->cdev->dev));
+	/* Print discipline string. */
+	if (device->discipline != NULL)
+		seq_printf(m, "(%s)", device->discipline->name);
+	else
+		seq_printf(m, "(none)");
+	/* Print kdev. */
+	if (block->gdp)
+		seq_printf(m, " at (%3d:%6d)",
+			   MAJOR(disk_devt(block->gdp)),
+			   MINOR(disk_devt(block->gdp)));
+	else
+		seq_printf(m, "  at (???:??????)");
+	/* Print device name. */
+	if (block->gdp)
+		seq_printf(m, " is %-8s", block->gdp->disk_name);
+	else
+		seq_printf(m, " is ????????");
+	/* Print devices features. */
+	substr = (device->features & DASD_FEATURE_READONLY) ? "(ro)" : " ";
+	seq_printf(m, "%4s: ", substr);
+	/* Print device status information. */
+	switch (device->state) {
+	case DASD_STATE_NEW:
+		seq_printf(m, "new");
+		break;
+	case DASD_STATE_KNOWN:
+		seq_printf(m, "detected");
+		break;
+	case DASD_STATE_BASIC:
+		seq_printf(m, "basic");
+		break;
+	case DASD_STATE_UNFMT:
+		seq_printf(m, "unformatted");
+		break;
+	case DASD_STATE_READY:
+	case DASD_STATE_ONLINE:
+		seq_printf(m, "active ");
+		if (dasd_check_blocksize(block->bp_block))
+			seq_printf(m, "n/f	 ");
+		else
+			seq_printf(m,
+				   "at blocksize: %d, %lld blocks, %lld MB",
+				   block->bp_block, block->blocks,
+				   ((block->bp_block >> 9) *
+				    block->blocks) >> 11);
+		break;
+	default:
+		seq_printf(m, "no stat");
+		break;
+	}
+	dasd_put_device(device);
+	if (dasd_probeonly)
+		seq_printf(m, "(probeonly)");
+	seq_printf(m, "\n");
+	return 0;
+}
+
+static void *dasd_devices_start(struct seq_file *m, loff_t *pos)
+{
+	if (*pos >= dasd_max_devindex)
+		return NULL;
+	return (void *)((unsigned long) *pos + 1);
+}
+
+static void *dasd_devices_next(struct seq_file *m, void *v, loff_t *pos)
+{
+	++*pos;
+	return dasd_devices_start(m, pos);
+}
+
+static void dasd_devices_stop(struct seq_file *m, void *v)
+{
+}
+
+static const struct seq_operations dasd_devices_seq_ops = {
+	.start		= dasd_devices_start,
+	.next		= dasd_devices_next,
+	.stop		= dasd_devices_stop,
+	.show		= dasd_devices_show,
+};
+
+static int dasd_devices_open(struct inode *inode, struct file *file)
+{
+	return seq_open(file, &dasd_devices_seq_ops);
+}
+
+static const struct file_operations dasd_devices_file_ops = {
+	.owner		= THIS_MODULE,
+	.open		= dasd_devices_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= seq_release,
+};
+
+#ifdef CONFIG_DASD_PROFILE
+static int dasd_stats_all_block_on(void)
+{
+	int i, rc;
+	struct dasd_device *device;
+
+	rc = 0;
+	for (i = 0; i < dasd_max_devindex; ++i) {
+		device = dasd_device_from_devindex(i);
+		if (IS_ERR(device))
+			continue;
+		if (device->block)
+			rc = dasd_profile_on(&device->block->profile);
+		dasd_put_device(device);
+		if (rc)
+			return rc;
+	}
+	return 0;
+}
+
+static void dasd_stats_all_block_off(void)
+{
+	int i;
+	struct dasd_device *device;
+
+	for (i = 0; i < dasd_max_devindex; ++i) {
+		device = dasd_device_from_devindex(i);
+		if (IS_ERR(device))
+			continue;
+		if (device->block)
+			dasd_profile_off(&device->block->profile);
+		dasd_put_device(device);
+	}
+}
+
+static void dasd_stats_all_block_reset(void)
+{
+	int i;
+	struct dasd_device *device;
+
+	for (i = 0; i < dasd_max_devindex; ++i) {
+		device = dasd_device_from_devindex(i);
+		if (IS_ERR(device))
+			continue;
+		if (device->block)
+			dasd_profile_reset(&device->block->profile);
+		dasd_put_device(device);
+	}
+}
+
+static void dasd_statistics_array(struct seq_file *m, unsigned int *array, int factor)
+{
+	int i;
+
+	for (i = 0; i < 32; i++) {
+		seq_printf(m, "%7d ", array[i] / factor);
+		if (i == 15)
+			seq_putc(m, '\n');
+	}
+	seq_putc(m, '\n');
+}
+#endif /* CONFIG_DASD_PROFILE */
+
+static int dasd_stats_proc_show(struct seq_file *m, void *v)
+{
+#ifdef CONFIG_DASD_PROFILE
+	struct dasd_profile_info *prof;
+	int factor;
+
+	/* check for active profiling */
+	if (!dasd_global_profile_level) {
+		seq_printf(m, "Statistics are off - they might be "
+				    "switched on using 'echo set on > "
+				    "/proc/dasd/statistics'\n");
+		return 0;
+	}
+	prof = &dasd_global_profile_data;
+
+	/* prevent counter 'overflow' on output */
+	for (factor = 1; (prof->dasd_io_reqs / factor) > 9999999;
+	     factor *= 10);
+
+	seq_printf(m, "%d dasd I/O requests\n", prof->dasd_io_reqs);
+	seq_printf(m, "with %u sectors(512B each)\n",
+		       prof->dasd_io_sects);
+	seq_printf(m, "Scale Factor is  %d\n", factor);
+	seq_printf(m,
+		       "   __<4	   ___8	   __16	   __32	   __64	   _128	"
+		       "   _256	   _512	   __1k	   __2k	   __4k	   __8k	"
+		       "   _16k	   _32k	   _64k	   128k\n");
+	seq_printf(m,
+		       "   _256	   _512	   __1M	   __2M	   __4M	   __8M	"
+		       "   _16M	   _32M	   _64M	   128M	   256M	   512M	"
+		       "   __1G	   __2G	   __4G " "   _>4G\n");
+
+	seq_printf(m, "Histogram of sizes (512B secs)\n");
+	dasd_statistics_array(m, prof->dasd_io_secs, factor);
+	seq_printf(m, "Histogram of I/O times (microseconds)\n");
+	dasd_statistics_array(m, prof->dasd_io_times, factor);
+	seq_printf(m, "Histogram of I/O times per sector\n");
+	dasd_statistics_array(m, prof->dasd_io_timps, factor);
+	seq_printf(m, "Histogram of I/O time till ssch\n");
+	dasd_statistics_array(m, prof->dasd_io_time1, factor);
+	seq_printf(m, "Histogram of I/O time between ssch and irq\n");
+	dasd_statistics_array(m, prof->dasd_io_time2, factor);
+	seq_printf(m, "Histogram of I/O time between ssch "
+			    "and irq per sector\n");
+	dasd_statistics_array(m, prof->dasd_io_time2ps, factor);
+	seq_printf(m, "Histogram of I/O time between irq and end\n");
+	dasd_statistics_array(m, prof->dasd_io_time3, factor);
+	seq_printf(m, "# of req in chanq at enqueuing (1..32) \n");
+	dasd_statistics_array(m, prof->dasd_io_nr_req, factor);
+#else
+	seq_printf(m, "Statistics are not activated in this kernel\n");
+#endif
+	return 0;
+}
+
+static int dasd_stats_proc_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, dasd_stats_proc_show, NULL);
+}
+
+static ssize_t dasd_stats_proc_write(struct file *file,
+		const char __user *user_buf, size_t user_len, loff_t *pos)
+{
+#ifdef CONFIG_DASD_PROFILE
+	char *buffer, *str;
+	int rc;
+
+	if (user_len > 65536)
+		user_len = 65536;
+	buffer = dasd_get_user_string(user_buf, user_len);
+	if (IS_ERR(buffer))
+		return PTR_ERR(buffer);
+
+	/* check for valid verbs */
+	str = skip_spaces(buffer);
+	if (strncmp(str, "set", 3) == 0 && isspace(str[3])) {
+		/* 'set xxx' was given */
+		str = skip_spaces(str + 4);
+		if (strcmp(str, "on") == 0) {
+			/* switch on statistics profiling */
+			rc = dasd_stats_all_block_on();
+			if (rc) {
+				dasd_stats_all_block_off();
+				goto out_error;
+			}
+			dasd_global_profile_reset();
+			dasd_global_profile_level = DASD_PROFILE_ON;
+			pr_info("The statistics feature has been switched "
+				"on\n");
+		} else if (strcmp(str, "off") == 0) {
+			/* switch off and reset statistics profiling */
+			dasd_global_profile_level = DASD_PROFILE_OFF;
+			dasd_global_profile_reset();
+			dasd_stats_all_block_off();
+			pr_info("The statistics feature has been switched "
+				"off\n");
+		} else
+			goto out_parse_error;
+	} else if (strncmp(str, "reset", 5) == 0) {
+		/* reset the statistics */
+		dasd_global_profile_reset();
+		dasd_stats_all_block_reset();
+		pr_info("The statistics have been reset\n");
+	} else
+		goto out_parse_error;
+	vfree(buffer);
+	return user_len;
+out_parse_error:
+	rc = -EINVAL;
+	pr_warning("%s is not a supported value for /proc/dasd/statistics\n",
+		str);
+out_error:
+	vfree(buffer);
+	return rc;
+#else
+	pr_warning("/proc/dasd/statistics: is not activated in this kernel\n");
+	return user_len;
+#endif				/* CONFIG_DASD_PROFILE */
+}
+
+static const struct file_operations dasd_stats_proc_fops = {
+	.owner		= THIS_MODULE,
+	.open		= dasd_stats_proc_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+	.write		= dasd_stats_proc_write,
+};
+
+/*
+ * Create dasd proc-fs entries.
+ * In case creation failed, cleanup and return -ENOENT.
+ */
+int
+dasd_proc_init(void)
+{
+	dasd_proc_root_entry = proc_mkdir("dasd", NULL);
+	if (!dasd_proc_root_entry)
+		goto out_nodasd;
+	dasd_devices_entry = proc_create("devices",
+					 S_IFREG | S_IRUGO | S_IWUSR,
+					 dasd_proc_root_entry,
+					 &dasd_devices_file_ops);
+	if (!dasd_devices_entry)
+		goto out_nodevices;
+	dasd_statistics_entry = proc_create("statistics",
+					    S_IFREG | S_IRUGO | S_IWUSR,
+					    dasd_proc_root_entry,
+					    &dasd_stats_proc_fops);
+	if (!dasd_statistics_entry)
+		goto out_nostatistics;
+	return 0;
+
+ out_nostatistics:
+	remove_proc_entry("devices", dasd_proc_root_entry);
+ out_nodevices:
+	remove_proc_entry("dasd", NULL);
+ out_nodasd:
+	return -ENOENT;
+}
+
+void
+dasd_proc_exit(void)
+{
+	remove_proc_entry("devices", dasd_proc_root_entry);
+	remove_proc_entry("statistics", dasd_proc_root_entry);
+	remove_proc_entry("dasd", NULL);
+}

drivers/s390/block/scm_blk.c

@@ -0,0 +1,494 @@
+/*
+ * Block driver for s390 storage class memory.
+ *
+ * Copyright IBM Corp. 2012
+ * Author(s): Sebastian Ott <sebott@linux.vnet.ibm.com>
+ */
+
+#define KMSG_COMPONENT "scm_block"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/interrupt.h>
+#include <linux/spinlock.h>
+#include <linux/module.h>
+#include <linux/blkdev.h>
+#include <linux/genhd.h>
+#include <linux/slab.h>
+#include <linux/list.h>
+#include <asm/eadm.h>
+#include "scm_blk.h"
+
+debug_info_t *scm_debug;
+static int scm_major;
+static DEFINE_SPINLOCK(list_lock);
+static LIST_HEAD(inactive_requests);
+static unsigned int nr_requests = 64;
+static atomic_t nr_devices = ATOMIC_INIT(0);
+module_param(nr_requests, uint, S_IRUGO);
+MODULE_PARM_DESC(nr_requests, "Number of parallel requests.");
+
+MODULE_DESCRIPTION("Block driver for s390 storage class memory.");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("scm:scmdev*");
+
+static void __scm_free_rq(struct scm_request *scmrq)
+{
+	struct aob_rq_header *aobrq = to_aobrq(scmrq);
+
+	free_page((unsigned long) scmrq->aob);
+	free_page((unsigned long) scmrq->aidaw);
+	__scm_free_rq_cluster(scmrq);
+	kfree(aobrq);
+}
+
+static void scm_free_rqs(void)
+{
+	struct list_head *iter, *safe;
+	struct scm_request *scmrq;
+
+	spin_lock_irq(&list_lock);
+	list_for_each_safe(iter, safe, &inactive_requests) {
+		scmrq = list_entry(iter, struct scm_request, list);
+		list_del(&scmrq->list);
+		__scm_free_rq(scmrq);
+	}
+	spin_unlock_irq(&list_lock);
+}
+
+static int __scm_alloc_rq(void)
+{
+	struct aob_rq_header *aobrq;
+	struct scm_request *scmrq;
+
+	aobrq = kzalloc(sizeof(*aobrq) + sizeof(*scmrq), GFP_KERNEL);
+	if (!aobrq)
+		return -ENOMEM;
+
+	scmrq = (void *) aobrq->data;
+	scmrq->aidaw = (void *) get_zeroed_page(GFP_DMA);
+	scmrq->aob = (void *) get_zeroed_page(GFP_DMA);
+	if (!scmrq->aob || !scmrq->aidaw) {
+		__scm_free_rq(scmrq);
+		return -ENOMEM;
+	}
+
+	if (__scm_alloc_rq_cluster(scmrq)) {
+		__scm_free_rq(scmrq);
+		return -ENOMEM;
+	}
+
+	INIT_LIST_HEAD(&scmrq->list);
+	spin_lock_irq(&list_lock);
+	list_add(&scmrq->list, &inactive_requests);
+	spin_unlock_irq(&list_lock);
+
+	return 0;
+}
+
+static int scm_alloc_rqs(unsigned int nrqs)
+{
+	int ret = 0;
+
+	while (nrqs-- && !ret)
+		ret = __scm_alloc_rq();
+
+	return ret;
+}
+
+static struct scm_request *scm_request_fetch(void)
+{
+	struct scm_request *scmrq = NULL;
+
+	spin_lock(&list_lock);
+	if (list_empty(&inactive_requests))
+		goto out;
+	scmrq = list_first_entry(&inactive_requests, struct scm_request, list);
+	list_del(&scmrq->list);
+out:
+	spin_unlock(&list_lock);
+	return scmrq;
+}
+
+static void scm_request_done(struct scm_request *scmrq)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&list_lock, flags);
+	list_add(&scmrq->list, &inactive_requests);
+	spin_unlock_irqrestore(&list_lock, flags);
+}
+
+static bool scm_permit_request(struct scm_blk_dev *bdev, struct request *req)
+{
+	return rq_data_dir(req) != WRITE || bdev->state != SCM_WR_PROHIBIT;
+}
+
+static void scm_request_prepare(struct scm_request *scmrq)
+{
+	struct scm_blk_dev *bdev = scmrq->bdev;
+	struct scm_device *scmdev = bdev->gendisk->private_data;
+	struct aidaw *aidaw = scmrq->aidaw;
+	struct msb *msb = &scmrq->aob->msb[0];
+	struct req_iterator iter;
+	struct bio_vec bv;
+
+	msb->bs = MSB_BS_4K;
+	scmrq->aob->request.msb_count = 1;
+	msb->scm_addr = scmdev->address +
+		((u64) blk_rq_pos(scmrq->request) << 9);
+	msb->oc = (rq_data_dir(scmrq->request) == READ) ?
+		MSB_OC_READ : MSB_OC_WRITE;
+	msb->flags |= MSB_FLAG_IDA;
+	msb->data_addr = (u64) aidaw;
+
+	rq_for_each_segment(bv, scmrq->request, iter) {
+		WARN_ON(bv.bv_offset);
+		msb->blk_count += bv.bv_len >> 12;
+		aidaw->data_addr = (u64) page_address(bv.bv_page);
+		aidaw++;
+	}
+}
+
+static inline void scm_request_init(struct scm_blk_dev *bdev,
+				    struct scm_request *scmrq,
+				    struct request *req)
+{
+	struct aob_rq_header *aobrq = to_aobrq(scmrq);
+	struct aob *aob = scmrq->aob;
+
+	memset(aob, 0, sizeof(*aob));
+	memset(scmrq->aidaw, 0, PAGE_SIZE);
+	aobrq->scmdev = bdev->scmdev;
+	aob->request.cmd_code = ARQB_CMD_MOVE;
+	aob->request.data = (u64) aobrq;
+	scmrq->request = req;
+	scmrq->bdev = bdev;
+	scmrq->retries = 4;
+	scmrq->error = 0;
+	scm_request_cluster_init(scmrq);
+}
+
+static void scm_ensure_queue_restart(struct scm_blk_dev *bdev)
+{
+	if (atomic_read(&bdev->queued_reqs)) {
+		/* Queue restart is triggered by the next interrupt. */
+		return;
+	}
+	blk_delay_queue(bdev->rq, SCM_QUEUE_DELAY);
+}
+
+void scm_request_requeue(struct scm_request *scmrq)
+{
+	struct scm_blk_dev *bdev = scmrq->bdev;
+
+	scm_release_cluster(scmrq);
+	blk_requeue_request(bdev->rq, scmrq->request);
+	atomic_dec(&bdev->queued_reqs);
+	scm_request_done(scmrq);
+	scm_ensure_queue_restart(bdev);
+}
+
+void scm_request_finish(struct scm_request *scmrq)
+{
+	struct scm_blk_dev *bdev = scmrq->bdev;
+
+	scm_release_cluster(scmrq);
+	blk_end_request_all(scmrq->request, scmrq->error);
+	atomic_dec(&bdev->queued_reqs);
+	scm_request_done(scmrq);
+}
+
+static void scm_blk_request(struct request_queue *rq)
+{
+	struct scm_device *scmdev = rq->queuedata;
+	struct scm_blk_dev *bdev = dev_get_drvdata(&scmdev->dev);
+	struct scm_request *scmrq;
+	struct request *req;
+	int ret;
+
+	while ((req = blk_peek_request(rq))) {
+		if (req->cmd_type != REQ_TYPE_FS) {
+			blk_start_request(req);
+			blk_dump_rq_flags(req, KMSG_COMPONENT " bad request");
+			blk_end_request_all(req, -EIO);
+			continue;
+		}
+
+		if (!scm_permit_request(bdev, req)) {
+			scm_ensure_queue_restart(bdev);
+			return;
+		}
+		scmrq = scm_request_fetch();
+		if (!scmrq) {
+			SCM_LOG(5, "no request");
+			scm_ensure_queue_restart(bdev);
+			return;
+		}
+		scm_request_init(bdev, scmrq, req);
+		if (!scm_reserve_cluster(scmrq)) {
+			SCM_LOG(5, "cluster busy");
+			scm_request_done(scmrq);
+			return;
+		}
+		if (scm_need_cluster_request(scmrq)) {
+			atomic_inc(&bdev->queued_reqs);
+			blk_start_request(req);
+			scm_initiate_cluster_request(scmrq);
+			return;
+		}
+		scm_request_prepare(scmrq);
+		atomic_inc(&bdev->queued_reqs);
+		blk_start_request(req);
+
+		ret = eadm_start_aob(scmrq->aob);
+		if (ret) {
+			SCM_LOG(5, "no subchannel");
+			scm_request_requeue(scmrq);
+			return;
+		}
+	}
+}
+
+static void __scmrq_log_error(struct scm_request *scmrq)
+{
+	struct aob *aob = scmrq->aob;
+
+	if (scmrq->error == -ETIMEDOUT)
+		SCM_LOG(1, "Request timeout");
+	else {
+		SCM_LOG(1, "Request error");
+		SCM_LOG_HEX(1, &aob->response, sizeof(aob->response));
+	}
+	if (scmrq->retries)
+		SCM_LOG(1, "Retry request");
+	else
+		pr_err("An I/O operation to SCM failed with rc=%d\n",
+		       scmrq->error);
+}
+
+void scm_blk_irq(struct scm_device *scmdev, void *data, int error)
+{
+	struct scm_request *scmrq = data;
+	struct scm_blk_dev *bdev = scmrq->bdev;
+
+	scmrq->error = error;
+	if (error)
+		__scmrq_log_error(scmrq);
+
+	spin_lock(&bdev->lock);
+	list_add_tail(&scmrq->list, &bdev->finished_requests);
+	spin_unlock(&bdev->lock);
+	tasklet_hi_schedule(&bdev->tasklet);
+}
+
+static void scm_blk_handle_error(struct scm_request *scmrq)
+{
+	struct scm_blk_dev *bdev = scmrq->bdev;
+	unsigned long flags;
+
+	if (scmrq->error != -EIO)
+		goto restart;
+
+	/* For -EIO the response block is valid. */
+	switch (scmrq->aob->response.eqc) {
+	case EQC_WR_PROHIBIT:
+		spin_lock_irqsave(&bdev->lock, flags);
+		if (bdev->state != SCM_WR_PROHIBIT)
+			pr_info("%lx: Write access to the SCM increment is suspended\n",
+				(unsigned long) bdev->scmdev->address);
+		bdev->state = SCM_WR_PROHIBIT;
+		spin_unlock_irqrestore(&bdev->lock, flags);
+		goto requeue;
+	default:
+		break;
+	}
+
+restart:
+	if (!eadm_start_aob(scmrq->aob))
+		return;
+
+requeue:
+	spin_lock_irqsave(&bdev->rq_lock, flags);
+	scm_request_requeue(scmrq);
+	spin_unlock_irqrestore(&bdev->rq_lock, flags);
+}
+
+static void scm_blk_tasklet(struct scm_blk_dev *bdev)
+{
+	struct scm_request *scmrq;
+	unsigned long flags;
+
+	spin_lock_irqsave(&bdev->lock, flags);
+	while (!list_empty(&bdev->finished_requests)) {
+		scmrq = list_first_entry(&bdev->finished_requests,
+					 struct scm_request, list);
+		list_del(&scmrq->list);
+		spin_unlock_irqrestore(&bdev->lock, flags);
+
+		if (scmrq->error && scmrq->retries-- > 0) {
+			scm_blk_handle_error(scmrq);
+
+			/* Request restarted or requeued, handle next. */
+			spin_lock_irqsave(&bdev->lock, flags);
+			continue;
+		}
+
+		if (scm_test_cluster_request(scmrq)) {
+			scm_cluster_request_irq(scmrq);
+			spin_lock_irqsave(&bdev->lock, flags);
+			continue;
+		}
+
+		scm_request_finish(scmrq);
+		spin_lock_irqsave(&bdev->lock, flags);
+	}
+	spin_unlock_irqrestore(&bdev->lock, flags);
+	/* Look out for more requests. */
+	blk_run_queue(bdev->rq);
+}
+
+static const struct block_device_operations scm_blk_devops = {
+	.owner = THIS_MODULE,
+};
+
+int scm_blk_dev_setup(struct scm_blk_dev *bdev, struct scm_device *scmdev)
+{
+	struct request_queue *rq;
+	int len, ret = -ENOMEM;
+	unsigned int devindex, nr_max_blk;
+
+	devindex = atomic_inc_return(&nr_devices) - 1;
+	/* scma..scmz + scmaa..scmzz */
+	if (devindex > 701) {
+		ret = -ENODEV;
+		goto out;
+	}
+
+	bdev->scmdev = scmdev;
+	bdev->state = SCM_OPER;
+	spin_lock_init(&bdev->rq_lock);
+	spin_lock_init(&bdev->lock);
+	INIT_LIST_HEAD(&bdev->finished_requests);
+	atomic_set(&bdev->queued_reqs, 0);
+	tasklet_init(&bdev->tasklet,
+		     (void (*)(unsigned long)) scm_blk_tasklet,
+		     (unsigned long) bdev);
+
+	rq = blk_init_queue(scm_blk_request, &bdev->rq_lock);
+	if (!rq)
+		goto out;
+
+	bdev->rq = rq;
+	nr_max_blk = min(scmdev->nr_max_block,
+			 (unsigned int) (PAGE_SIZE / sizeof(struct aidaw)));
+
+	blk_queue_logical_block_size(rq, 1 << 12);
+	blk_queue_max_hw_sectors(rq, nr_max_blk << 3); /* 8 * 512 = blk_size */
+	blk_queue_max_segments(rq, nr_max_blk);
+	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, rq);
+	queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, rq);
+	scm_blk_dev_cluster_setup(bdev);
+
+	bdev->gendisk = alloc_disk(SCM_NR_PARTS);
+	if (!bdev->gendisk)
+		goto out_queue;
+
+	rq->queuedata = scmdev;
+	bdev->gendisk->driverfs_dev = &scmdev->dev;
+	bdev->gendisk->private_data = scmdev;
+	bdev->gendisk->fops = &scm_blk_devops;
+	bdev->gendisk->queue = rq;
+	bdev->gendisk->major = scm_major;
+	bdev->gendisk->first_minor = devindex * SCM_NR_PARTS;
+
+	len = snprintf(bdev->gendisk->disk_name, DISK_NAME_LEN, "scm");
+	if (devindex > 25) {
+		len += snprintf(bdev->gendisk->disk_name + len,
+				DISK_NAME_LEN - len, "%c",
+				'a' + (devindex / 26) - 1);
+		devindex = devindex % 26;
+	}
+	snprintf(bdev->gendisk->disk_name + len, DISK_NAME_LEN - len, "%c",
+		 'a' + devindex);
+
+	/* 512 byte sectors */
+	set_capacity(bdev->gendisk, scmdev->size >> 9);
+	add_disk(bdev->gendisk);
+	return 0;
+
+out_queue:
+	blk_cleanup_queue(rq);
+out:
+	atomic_dec(&nr_devices);
+	return ret;
+}
+
+void scm_blk_dev_cleanup(struct scm_blk_dev *bdev)
+{
+	tasklet_kill(&bdev->tasklet);
+	del_gendisk(bdev->gendisk);
+	blk_cleanup_queue(bdev->gendisk->queue);
+	put_disk(bdev->gendisk);
+}
+
+void scm_blk_set_available(struct scm_blk_dev *bdev)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&bdev->lock, flags);
+	if (bdev->state == SCM_WR_PROHIBIT)
+		pr_info("%lx: Write access to the SCM increment is restored\n",
+			(unsigned long) bdev->scmdev->address);
+	bdev->state = SCM_OPER;
+	spin_unlock_irqrestore(&bdev->lock, flags);
+}
+
+static int __init scm_blk_init(void)
+{
+	int ret = -EINVAL;
+
+	if (!scm_cluster_size_valid())
+		goto out;
+
+	ret = register_blkdev(0, "scm");
+	if (ret < 0)
+		goto out;
+
+	scm_major = ret;
+	ret = scm_alloc_rqs(nr_requests);
+	if (ret)
+		goto out_free;
+
+	scm_debug = debug_register("scm_log", 16, 1, 16);
+	if (!scm_debug) {
+		ret = -ENOMEM;
+		goto out_free;
+	}
+
+	debug_register_view(scm_debug, &debug_hex_ascii_view);
+	debug_set_level(scm_debug, 2);
+
+	ret = scm_drv_init();
+	if (ret)
+		goto out_dbf;
+
+	return ret;
+
+out_dbf:
+	debug_unregister(scm_debug);
+out_free:
+	scm_free_rqs();
+	unregister_blkdev(scm_major, "scm");
+out:
+	return ret;
+}
+module_init(scm_blk_init);
+
+static void __exit scm_blk_cleanup(void)
+{
+	scm_drv_cleanup();
+	debug_unregister(scm_debug);
+	scm_free_rqs();
+	unregister_blkdev(scm_major, "scm");
+}
+module_exit(scm_blk_cleanup);

drivers/s390/block/scm_blk.h

@@ -0,0 +1,134 @@
+#ifndef SCM_BLK_H
+#define SCM_BLK_H
+
+#include <linux/interrupt.h>
+#include <linux/spinlock.h>
+#include <linux/blkdev.h>
+#include <linux/genhd.h>
+#include <linux/list.h>
+
+#include <asm/debug.h>
+#include <asm/eadm.h>
+
+#define SCM_NR_PARTS 8
+#define SCM_QUEUE_DELAY 5
+
+struct scm_blk_dev {
+	struct tasklet_struct tasklet;
+	struct request_queue *rq;
+	struct gendisk *gendisk;
+	struct scm_device *scmdev;
+	spinlock_t rq_lock;	/* guard the request queue */
+	spinlock_t lock;	/* guard the rest of the blockdev */
+	atomic_t queued_reqs;
+	enum {SCM_OPER, SCM_WR_PROHIBIT} state;
+	struct list_head finished_requests;
+#ifdef CONFIG_SCM_BLOCK_CLUSTER_WRITE
+	struct list_head cluster_list;
+#endif
+};
+
+struct scm_request {
+	struct scm_blk_dev *bdev;
+	struct request *request;
+	struct aidaw *aidaw;
+	struct aob *aob;
+	struct list_head list;
+	u8 retries;
+	int error;
+#ifdef CONFIG_SCM_BLOCK_CLUSTER_WRITE
+	struct {
+		enum {CLUSTER_NONE, CLUSTER_READ, CLUSTER_WRITE} state;
+		struct list_head list;
+		void **buf;
+	} cluster;
+#endif
+};
+
+#define to_aobrq(rq) container_of((void *) rq, struct aob_rq_header, data)
+
+int scm_blk_dev_setup(struct scm_blk_dev *, struct scm_device *);
+void scm_blk_dev_cleanup(struct scm_blk_dev *);
+void scm_blk_set_available(struct scm_blk_dev *);
+void scm_blk_irq(struct scm_device *, void *, int);
+
+void scm_request_finish(struct scm_request *);
+void scm_request_requeue(struct scm_request *);
+
+int scm_drv_init(void);
+void scm_drv_cleanup(void);
+
+#ifdef CONFIG_SCM_BLOCK_CLUSTER_WRITE
+void __scm_free_rq_cluster(struct scm_request *);
+int __scm_alloc_rq_cluster(struct scm_request *);
+void scm_request_cluster_init(struct scm_request *);
+bool scm_reserve_cluster(struct scm_request *);
+void scm_release_cluster(struct scm_request *);
+void scm_blk_dev_cluster_setup(struct scm_blk_dev *);
+bool scm_need_cluster_request(struct scm_request *);
+void scm_initiate_cluster_request(struct scm_request *);
+void scm_cluster_request_irq(struct scm_request *);
+bool scm_test_cluster_request(struct scm_request *);
+bool scm_cluster_size_valid(void);
+#else /* CONFIG_SCM_BLOCK_CLUSTER_WRITE */
+static inline void __scm_free_rq_cluster(struct scm_request *scmrq) {}
+static inline int __scm_alloc_rq_cluster(struct scm_request *scmrq)
+{
+	return 0;
+}
+static inline void scm_request_cluster_init(struct scm_request *scmrq) {}
+static inline bool scm_reserve_cluster(struct scm_request *scmrq)
+{
+	return true;
+}
+static inline void scm_release_cluster(struct scm_request *scmrq) {}
+static inline void scm_blk_dev_cluster_setup(struct scm_blk_dev *bdev) {}
+static inline bool scm_need_cluster_request(struct scm_request *scmrq)
+{
+	return false;
+}
+static inline void scm_initiate_cluster_request(struct scm_request *scmrq) {}
+static inline void scm_cluster_request_irq(struct scm_request *scmrq) {}
+static inline bool scm_test_cluster_request(struct scm_request *scmrq)
+{
+	return false;
+}
+static inline bool scm_cluster_size_valid(void)
+{
+	return true;
+}
+#endif /* CONFIG_SCM_BLOCK_CLUSTER_WRITE */
+
+extern debug_info_t *scm_debug;
+
+#define SCM_LOG(imp, txt) do {					\
+		debug_text_event(scm_debug, imp, txt);		\
+	} while (0)
+
+static inline void SCM_LOG_HEX(int level, void *data, int length)
+{
+	if (!debug_level_enabled(scm_debug, level))
+		return;
+	while (length > 0) {
+		debug_event(scm_debug, level, data, length);
+		length -= scm_debug->buf_size;
+		data += scm_debug->buf_size;
+	}
+}
+
+static inline void SCM_LOG_STATE(int level, struct scm_device *scmdev)
+{
+	struct {
+		u64 address;
+		u8 oper_state;
+		u8 rank;
+	} __packed data = {
+		.address = scmdev->address,
+		.oper_state = scmdev->attrs.oper_state,
+		.rank = scmdev->attrs.rank,
+	};
+
+	SCM_LOG_HEX(level, &data, sizeof(data));
+}
+
+#endif /* SCM_BLK_H */

drivers/s390/block/scm_blk_cluster.c

@@ -0,0 +1,230 @@
+/*
+ * Block driver for s390 storage class memory.
+ *
+ * Copyright IBM Corp. 2012
+ * Author(s): Sebastian Ott <sebott@linux.vnet.ibm.com>
+ */
+
+#include <linux/spinlock.h>
+#include <linux/module.h>
+#include <linux/blkdev.h>
+#include <linux/genhd.h>
+#include <linux/slab.h>
+#include <linux/list.h>
+#include <asm/eadm.h>
+#include "scm_blk.h"
+
+static unsigned int write_cluster_size = 64;
+module_param(write_cluster_size, uint, S_IRUGO);
+MODULE_PARM_DESC(write_cluster_size,
+		 "Number of pages used for contiguous writes.");
+
+#define CLUSTER_SIZE (write_cluster_size * PAGE_SIZE)
+
+void __scm_free_rq_cluster(struct scm_request *scmrq)
+{
+	int i;
+
+	if (!scmrq->cluster.buf)
+		return;
+
+	for (i = 0; i < 2 * write_cluster_size; i++)
+		free_page((unsigned long) scmrq->cluster.buf[i]);
+
+	kfree(scmrq->cluster.buf);
+}
+
+int __scm_alloc_rq_cluster(struct scm_request *scmrq)
+{
+	int i;
+
+	scmrq->cluster.buf = kzalloc(sizeof(void *) * 2 * write_cluster_size,
+				 GFP_KERNEL);
+	if (!scmrq->cluster.buf)
+		return -ENOMEM;
+
+	for (i = 0; i < 2 * write_cluster_size; i++) {
+		scmrq->cluster.buf[i] = (void *) get_zeroed_page(GFP_DMA);
+		if (!scmrq->cluster.buf[i])
+			return -ENOMEM;
+	}
+	INIT_LIST_HEAD(&scmrq->cluster.list);
+	return 0;
+}
+
+void scm_request_cluster_init(struct scm_request *scmrq)
+{
+	scmrq->cluster.state = CLUSTER_NONE;
+}
+
+static bool clusters_intersect(struct scm_request *A, struct scm_request *B)
+{
+	unsigned long firstA, lastA, firstB, lastB;
+
+	firstA = ((u64) blk_rq_pos(A->request) << 9) / CLUSTER_SIZE;
+	lastA = (((u64) blk_rq_pos(A->request) << 9) +
+		    blk_rq_bytes(A->request) - 1) / CLUSTER_SIZE;
+
+	firstB = ((u64) blk_rq_pos(B->request) << 9) / CLUSTER_SIZE;
+	lastB = (((u64) blk_rq_pos(B->request) << 9) +
+		    blk_rq_bytes(B->request) - 1) / CLUSTER_SIZE;
+
+	return (firstB <= lastA && firstA <= lastB);
+}
+
+bool scm_reserve_cluster(struct scm_request *scmrq)
+{
+	struct scm_blk_dev *bdev = scmrq->bdev;
+	struct scm_request *iter;
+
+	if (write_cluster_size == 0)
+		return true;
+
+	spin_lock(&bdev->lock);
+	list_for_each_entry(iter, &bdev->cluster_list, cluster.list) {
+		if (clusters_intersect(scmrq, iter) &&
+		    (rq_data_dir(scmrq->request) == WRITE ||
+		     rq_data_dir(iter->request) == WRITE)) {
+			spin_unlock(&bdev->lock);
+			return false;
+		}
+	}
+	list_add(&scmrq->cluster.list, &bdev->cluster_list);
+	spin_unlock(&bdev->lock);
+
+	return true;
+}
+
+void scm_release_cluster(struct scm_request *scmrq)
+{
+	struct scm_blk_dev *bdev = scmrq->bdev;
+	unsigned long flags;
+
+	if (write_cluster_size == 0)
+		return;
+
+	spin_lock_irqsave(&bdev->lock, flags);
+	list_del(&scmrq->cluster.list);
+	spin_unlock_irqrestore(&bdev->lock, flags);
+}
+
+void scm_blk_dev_cluster_setup(struct scm_blk_dev *bdev)
+{
+	INIT_LIST_HEAD(&bdev->cluster_list);
+	blk_queue_io_opt(bdev->rq, CLUSTER_SIZE);
+}
+
+static void scm_prepare_cluster_request(struct scm_request *scmrq)
+{
+	struct scm_blk_dev *bdev = scmrq->bdev;
+	struct scm_device *scmdev = bdev->gendisk->private_data;
+	struct request *req = scmrq->request;
+	struct aidaw *aidaw = scmrq->aidaw;
+	struct msb *msb = &scmrq->aob->msb[0];
+	struct req_iterator iter;
+	struct bio_vec bv;
+	int i = 0;
+	u64 addr;
+
+	switch (scmrq->cluster.state) {
+	case CLUSTER_NONE:
+		scmrq->cluster.state = CLUSTER_READ;
+		/* fall through */
+	case CLUSTER_READ:
+		scmrq->aob->request.msb_count = 1;
+		msb->bs = MSB_BS_4K;
+		msb->oc = MSB_OC_READ;
+		msb->flags = MSB_FLAG_IDA;
+		msb->data_addr = (u64) aidaw;
+		msb->blk_count = write_cluster_size;
+
+		addr = scmdev->address + ((u64) blk_rq_pos(req) << 9);
+		msb->scm_addr = round_down(addr, CLUSTER_SIZE);
+
+		if (msb->scm_addr !=
+		    round_down(addr + (u64) blk_rq_bytes(req) - 1,
+			       CLUSTER_SIZE))
+			msb->blk_count = 2 * write_cluster_size;
+
+		for (i = 0; i < msb->blk_count; i++) {
+			aidaw->data_addr = (u64) scmrq->cluster.buf[i];
+			aidaw++;
+		}
+
+		break;
+	case CLUSTER_WRITE:
+		msb->oc = MSB_OC_WRITE;
+
+		for (addr = msb->scm_addr;
+		     addr < scmdev->address + ((u64) blk_rq_pos(req) << 9);
+		     addr += PAGE_SIZE) {
+			aidaw->data_addr = (u64) scmrq->cluster.buf[i];
+			aidaw++;
+			i++;
+		}
+		rq_for_each_segment(bv, req, iter) {
+			aidaw->data_addr = (u64) page_address(bv.bv_page);
+			aidaw++;
+			i++;
+		}
+		for (; i < msb->blk_count; i++) {
+			aidaw->data_addr = (u64) scmrq->cluster.buf[i];
+			aidaw++;
+		}
+		break;
+	}
+}
+
+bool scm_need_cluster_request(struct scm_request *scmrq)
+{
+	if (rq_data_dir(scmrq->request) == READ)
+		return false;
+
+	return blk_rq_bytes(scmrq->request) < CLUSTER_SIZE;
+}
+
+/* Called with queue lock held. */
+void scm_initiate_cluster_request(struct scm_request *scmrq)
+{
+	scm_prepare_cluster_request(scmrq);
+	if (eadm_start_aob(scmrq->aob))
+		scm_request_requeue(scmrq);
+}
+
+bool scm_test_cluster_request(struct scm_request *scmrq)
+{
+	return scmrq->cluster.state != CLUSTER_NONE;
+}
+
+void scm_cluster_request_irq(struct scm_request *scmrq)
+{
+	struct scm_blk_dev *bdev = scmrq->bdev;
+	unsigned long flags;
+
+	switch (scmrq->cluster.state) {
+	case CLUSTER_NONE:
+		BUG();
+		break;
+	case CLUSTER_READ:
+		if (scmrq->error) {
+			scm_request_finish(scmrq);
+			break;
+		}
+		scmrq->cluster.state = CLUSTER_WRITE;
+		spin_lock_irqsave(&bdev->rq_lock, flags);
+		scm_initiate_cluster_request(scmrq);
+		spin_unlock_irqrestore(&bdev->rq_lock, flags);
+		break;
+	case CLUSTER_WRITE:
+		scm_request_finish(scmrq);
+		break;
+	}
+}
+
+bool scm_cluster_size_valid(void)
+{
+	if (write_cluster_size == 1 || write_cluster_size > 128)
+		return false;
+
+	return !(write_cluster_size & (write_cluster_size - 1));
+}

drivers/s390/block/scm_drv.c

@@ -0,0 +1,92 @@
+/*
+ * Device driver for s390 storage class memory.
+ *
+ * Copyright IBM Corp. 2012
+ * Author(s): Sebastian Ott <sebott@linux.vnet.ibm.com>
+ */
+
+#define KMSG_COMPONENT "scm_block"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <asm/eadm.h>
+#include "scm_blk.h"
+
+static void scm_notify(struct scm_device *scmdev, enum scm_event event)
+{
+	struct scm_blk_dev *bdev = dev_get_drvdata(&scmdev->dev);
+
+	switch (event) {
+	case SCM_CHANGE:
+		pr_info("%lx: The capabilities of the SCM increment changed\n",
+			(unsigned long) scmdev->address);
+		SCM_LOG(2, "State changed");
+		SCM_LOG_STATE(2, scmdev);
+		break;
+	case SCM_AVAIL:
+		SCM_LOG(2, "Increment available");
+		SCM_LOG_STATE(2, scmdev);
+		scm_blk_set_available(bdev);
+		break;
+	}
+}
+
+static int scm_probe(struct scm_device *scmdev)
+{
+	struct scm_blk_dev *bdev;
+	int ret;
+
+	SCM_LOG(2, "probe");
+	SCM_LOG_STATE(2, scmdev);
+
+	if (scmdev->attrs.oper_state != OP_STATE_GOOD)
+		return -EINVAL;
+
+	bdev = kzalloc(sizeof(*bdev), GFP_KERNEL);
+	if (!bdev)
+		return -ENOMEM;
+
+	dev_set_drvdata(&scmdev->dev, bdev);
+	ret = scm_blk_dev_setup(bdev, scmdev);
+	if (ret) {
+		dev_set_drvdata(&scmdev->dev, NULL);
+		kfree(bdev);
+		goto out;
+	}
+
+out:
+	return ret;
+}
+
+static int scm_remove(struct scm_device *scmdev)
+{
+	struct scm_blk_dev *bdev = dev_get_drvdata(&scmdev->dev);
+
+	scm_blk_dev_cleanup(bdev);
+	dev_set_drvdata(&scmdev->dev, NULL);
+	kfree(bdev);
+
+	return 0;
+}
+
+static struct scm_driver scm_drv = {
+	.drv = {
+		.name = "scm_block",
+		.owner = THIS_MODULE,
+	},
+	.notify = scm_notify,
+	.probe = scm_probe,
+	.remove = scm_remove,
+	.handler = scm_blk_irq,
+};
+
+int __init scm_drv_init(void)
+{
+	return scm_driver_register(&scm_drv);
+}
+
+void scm_drv_cleanup(void)
+{
+	scm_driver_unregister(&scm_drv);
+}

drivers/s390/block/xpram.c

@@ -0,0 +1,479 @@
+/*
+ * Xpram.c -- the S/390 expanded memory RAM-disk
+ *           
+ * significant parts of this code are based on
+ * the sbull device driver presented in
+ * A. Rubini: Linux Device Drivers
+ *
+ * Author of XPRAM specific coding: Reinhard Buendgen
+ *                                  buendgen@de.ibm.com
+ * Rewrite for 2.5: Martin Schwidefsky <schwidefsky@de.ibm.com>
+ *
+ * External interfaces:
+ *   Interfaces to linux kernel
+ *        xpram_setup: read kernel parameters
+ *   Device specific file operations
+ *        xpram_iotcl
+ *        xpram_open
+ *
+ * "ad-hoc" partitioning:
+ *    the expanded memory can be partitioned among several devices 
+ *    (with different minors). The partitioning set up can be
+ *    set by kernel or module parameters (int devs & int sizes[])
+ *
+ * Potential future improvements:
+ *   generic hard disk support to replace ad-hoc partitioning
+ */
+
+#define KMSG_COMPONENT "xpram"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/ctype.h>  /* isdigit, isxdigit */
+#include <linux/errno.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/blkpg.h>
+#include <linux/hdreg.h>  /* HDIO_GETGEO */
+#include <linux/device.h>
+#include <linux/bio.h>
+#include <linux/suspend.h>
+#include <linux/platform_device.h>
+#include <linux/gfp.h>
+#include <asm/uaccess.h>
+
+#define XPRAM_NAME	"xpram"
+#define XPRAM_DEVS	1	/* one partition */
+#define XPRAM_MAX_DEVS	32	/* maximal number of devices (partitions) */
+
+typedef struct {
+	unsigned int	size;		/* size of xpram segment in pages */
+	unsigned int	offset;		/* start page of xpram segment */
+} xpram_device_t;
+
+static xpram_device_t xpram_devices[XPRAM_MAX_DEVS];
+static unsigned int xpram_sizes[XPRAM_MAX_DEVS];
+static struct gendisk *xpram_disks[XPRAM_MAX_DEVS];
+static struct request_queue *xpram_queues[XPRAM_MAX_DEVS];
+static unsigned int xpram_pages;
+static int xpram_devs;
+
+/*
+ * Parameter parsing functions.
+ */
+static int devs = XPRAM_DEVS;
+static char *sizes[XPRAM_MAX_DEVS];
+
+module_param(devs, int, 0);
+module_param_array(sizes, charp, NULL, 0);
+
+MODULE_PARM_DESC(devs, "number of devices (\"partitions\"), " \
+		 "the default is " __MODULE_STRING(XPRAM_DEVS) "\n");
+MODULE_PARM_DESC(sizes, "list of device (partition) sizes " \
+		 "the defaults are 0s \n" \
+		 "All devices with size 0 equally partition the "
+		 "remaining space on the expanded strorage not "
+		 "claimed by explicit sizes\n");
+MODULE_LICENSE("GPL");
+
+/*
+ * Copy expanded memory page (4kB) into main memory                  
+ * Arguments                                                         
+ *           page_addr:    address of target page                    
+ *           xpage_index:  index of expandeded memory page           
+ * Return value                                                      
+ *           0:            if operation succeeds
+ *           -EIO:         if pgin failed
+ *           -ENXIO:       if xpram has vanished
+ */
+static int xpram_page_in (unsigned long page_addr, unsigned int xpage_index)
+{
+	int cc = 2;	/* return unused cc 2 if pgin traps */
+
+	asm volatile(
+		"	.insn	rre,0xb22e0000,%1,%2\n"  /* pgin %1,%2 */
+		"0:	ipm	%0\n"
+		"	srl	%0,28\n"
+		"1:\n"
+		EX_TABLE(0b,1b)
+		: "+d" (cc) : "a" (__pa(page_addr)), "d" (xpage_index) : "cc");
+	if (cc == 3)
+		return -ENXIO;
+	if (cc == 2)
+		return -ENXIO;
+	if (cc == 1)
+		return -EIO;
+	return 0;
+}
+
+/*
+ * Copy a 4kB page of main memory to an expanded memory page          
+ * Arguments                                                          
+ *           page_addr:    address of source page                     
+ *           xpage_index:  index of expandeded memory page            
+ * Return value                                                       
+ *           0:            if operation succeeds
+ *           -EIO:         if pgout failed
+ *           -ENXIO:       if xpram has vanished
+ */
+static long xpram_page_out (unsigned long page_addr, unsigned int xpage_index)
+{
+	int cc = 2;	/* return unused cc 2 if pgin traps */
+
+	asm volatile(
+		"	.insn	rre,0xb22f0000,%1,%2\n"  /* pgout %1,%2 */
+		"0:	ipm	%0\n"
+		"	srl	%0,28\n"
+		"1:\n"
+		EX_TABLE(0b,1b)
+		: "+d" (cc) : "a" (__pa(page_addr)), "d" (xpage_index) : "cc");
+	if (cc == 3)
+		return -ENXIO;
+	if (cc == 2)
+		return -ENXIO;
+	if (cc == 1)
+		return -EIO;
+	return 0;
+}
+
+/*
+ * Check if xpram is available.
+ */
+static int xpram_present(void)
+{
+	unsigned long mem_page;
+	int rc;
+
+	mem_page = (unsigned long) __get_free_page(GFP_KERNEL);
+	if (!mem_page)
+		return -ENOMEM;
+	rc = xpram_page_in(mem_page, 0);
+	free_page(mem_page);
+	return rc ? -ENXIO : 0;
+}
+
+/*
+ * Return index of the last available xpram page.
+ */
+static unsigned long xpram_highest_page_index(void)
+{
+	unsigned int page_index, add_bit;
+	unsigned long mem_page;
+
+	mem_page = (unsigned long) __get_free_page(GFP_KERNEL);
+	if (!mem_page)
+		return 0;
+
+	page_index = 0;
+	add_bit = 1ULL << (sizeof(unsigned int)*8 - 1);
+	while (add_bit > 0) {
+		if (xpram_page_in(mem_page, page_index | add_bit) == 0)
+			page_index |= add_bit;
+		add_bit >>= 1;
+	}
+
+	free_page (mem_page);
+
+	return page_index;
+}
+
+/*
+ * Block device make request function.
+ */
+static void xpram_make_request(struct request_queue *q, struct bio *bio)
+{
+	xpram_device_t *xdev = bio->bi_bdev->bd_disk->private_data;
+	struct bio_vec bvec;
+	struct bvec_iter iter;
+	unsigned int index;
+	unsigned long page_addr;
+	unsigned long bytes;
+
+	if ((bio->bi_iter.bi_sector & 7) != 0 ||
+	    (bio->bi_iter.bi_size & 4095) != 0)
+		/* Request is not page-aligned. */
+		goto fail;
+	if ((bio->bi_iter.bi_size >> 12) > xdev->size)
+		/* Request size is no page-aligned. */
+		goto fail;
+	if ((bio->bi_iter.bi_sector >> 3) > 0xffffffffU - xdev->offset)
+		goto fail;
+	index = (bio->bi_iter.bi_sector >> 3) + xdev->offset;
+	bio_for_each_segment(bvec, bio, iter) {
+		page_addr = (unsigned long)
+			kmap(bvec.bv_page) + bvec.bv_offset;
+		bytes = bvec.bv_len;
+		if ((page_addr & 4095) != 0 || (bytes & 4095) != 0)
+			/* More paranoia. */
+			goto fail;
+		while (bytes > 0) {
+			if (bio_data_dir(bio) == READ) {
+				if (xpram_page_in(page_addr, index) != 0)
+					goto fail;
+			} else {
+				if (xpram_page_out(page_addr, index) != 0)
+					goto fail;
+			}
+			page_addr += 4096;
+			bytes -= 4096;
+			index++;
+		}
+	}
+	set_bit(BIO_UPTODATE, &bio->bi_flags);
+	bio_endio(bio, 0);
+	return;
+fail:
+	bio_io_error(bio);
+}
+
+static int xpram_getgeo(struct block_device *bdev, struct hd_geometry *geo)
+{
+	unsigned long size;
+
+	/*
+	 * get geometry: we have to fake one...  trim the size to a
+	 * multiple of 64 (32k): tell we have 16 sectors, 4 heads,
+	 * whatever cylinders. Tell also that data starts at sector. 4.
+	 */
+	size = (xpram_pages * 8) & ~0x3f;
+	geo->cylinders = size >> 6;
+	geo->heads = 4;
+	geo->sectors = 16;
+	geo->start = 4;
+	return 0;
+}
+
+static const struct block_device_operations xpram_devops =
+{
+	.owner	= THIS_MODULE,
+	.getgeo	= xpram_getgeo,
+};
+
+/*
+ * Setup xpram_sizes array.
+ */
+static int __init xpram_setup_sizes(unsigned long pages)
+{
+	unsigned long mem_needed;
+	unsigned long mem_auto;
+	unsigned long long size;
+	char *sizes_end;
+	int mem_auto_no;
+	int i;
+
+	/* Check number of devices. */
+	if (devs <= 0 || devs > XPRAM_MAX_DEVS) {
+		pr_err("%d is not a valid number of XPRAM devices\n",devs);
+		return -EINVAL;
+	}
+	xpram_devs = devs;
+
+	/*
+	 * Copy sizes array to xpram_sizes and align partition
+	 * sizes to page boundary.
+	 */
+	mem_needed = 0;
+	mem_auto_no = 0;
+	for (i = 0; i < xpram_devs; i++) {
+		if (sizes[i]) {
+			size = simple_strtoull(sizes[i], &sizes_end, 0);
+			switch (*sizes_end) {
+			case 'g':
+			case 'G':
+				size <<= 20;
+				break;
+			case 'm':
+			case 'M':
+				size <<= 10;
+			}
+			xpram_sizes[i] = (size + 3) & -4UL;
+		}
+		if (xpram_sizes[i])
+			mem_needed += xpram_sizes[i];
+		else
+			mem_auto_no++;
+	}
+	
+	pr_info("  number of devices (partitions): %d \n", xpram_devs);
+	for (i = 0; i < xpram_devs; i++) {
+		if (xpram_sizes[i])
+			pr_info("  size of partition %d: %u kB\n",
+				i, xpram_sizes[i]);
+		else
+			pr_info("  size of partition %d to be set "
+				"automatically\n",i);
+	}
+	pr_info("  memory needed (for sized partitions): %lu kB\n",
+		mem_needed);
+	pr_info("  partitions to be sized automatically: %d\n",
+		mem_auto_no);
+
+	if (mem_needed > pages * 4) {
+		pr_err("Not enough expanded memory available\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * partitioning:
+	 * xpram_sizes[i] != 0; partition i has size xpram_sizes[i] kB
+	 * else:             ; all partitions with zero xpram_sizes[i]
+	 *                     partition equally the remaining space
+	 */
+	if (mem_auto_no) {
+		mem_auto = ((pages - mem_needed / 4) / mem_auto_no) * 4;
+		pr_info("  automatically determined "
+			"partition size: %lu kB\n", mem_auto);
+		for (i = 0; i < xpram_devs; i++)
+			if (xpram_sizes[i] == 0)
+				xpram_sizes[i] = mem_auto;
+	}
+	return 0;
+}
+
+static int __init xpram_setup_blkdev(void)
+{
+	unsigned long offset;
+	int i, rc = -ENOMEM;
+
+	for (i = 0; i < xpram_devs; i++) {
+		xpram_disks[i] = alloc_disk(1);
+		if (!xpram_disks[i])
+			goto out;
+		xpram_queues[i] = blk_alloc_queue(GFP_KERNEL);
+		if (!xpram_queues[i]) {
+			put_disk(xpram_disks[i]);
+			goto out;
+		}
+		queue_flag_set_unlocked(QUEUE_FLAG_NONROT, xpram_queues[i]);
+		queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, xpram_queues[i]);
+		blk_queue_make_request(xpram_queues[i], xpram_make_request);
+		blk_queue_logical_block_size(xpram_queues[i], 4096);
+	}
+
+	/*
+	 * Register xpram major.
+	 */
+	rc = register_blkdev(XPRAM_MAJOR, XPRAM_NAME);
+	if (rc < 0)
+		goto out;
+
+	/*
+	 * Setup device structures.
+	 */
+	offset = 0;
+	for (i = 0; i < xpram_devs; i++) {
+		struct gendisk *disk = xpram_disks[i];
+
+		xpram_devices[i].size = xpram_sizes[i] / 4;
+		xpram_devices[i].offset = offset;
+		offset += xpram_devices[i].size;
+		disk->major = XPRAM_MAJOR;
+		disk->first_minor = i;
+		disk->fops = &xpram_devops;
+		disk->private_data = &xpram_devices[i];
+		disk->queue = xpram_queues[i];
+		sprintf(disk->disk_name, "slram%d", i);
+		set_capacity(disk, xpram_sizes[i] << 1);
+		add_disk(disk);
+	}
+
+	return 0;
+out:
+	while (i--) {
+		blk_cleanup_queue(xpram_queues[i]);
+		put_disk(xpram_disks[i]);
+	}
+	return rc;
+}
+
+/*
+ * Resume failed: Print error message and call panic.
+ */
+static void xpram_resume_error(const char *message)
+{
+	pr_err("Resuming the system failed: %s\n", message);
+	panic("xpram resume error\n");
+}
+
+/*
+ * Check if xpram setup changed between suspend and resume.
+ */
+static int xpram_restore(struct device *dev)
+{
+	if (!xpram_pages)
+		return 0;
+	if (xpram_present() != 0)
+		xpram_resume_error("xpram disappeared");
+	if (xpram_pages != xpram_highest_page_index() + 1)
+		xpram_resume_error("Size of xpram changed");
+	return 0;
+}
+
+static const struct dev_pm_ops xpram_pm_ops = {
+	.restore	= xpram_restore,
+};
+
+static struct platform_driver xpram_pdrv = {
+	.driver = {
+		.name	= XPRAM_NAME,
+		.owner	= THIS_MODULE,
+		.pm	= &xpram_pm_ops,
+	},
+};
+
+static struct platform_device *xpram_pdev;
+
+/*
+ * Finally, the init/exit functions.
+ */
+static void __exit xpram_exit(void)
+{
+	int i;
+	for (i = 0; i < xpram_devs; i++) {
+		del_gendisk(xpram_disks[i]);
+		blk_cleanup_queue(xpram_queues[i]);
+		put_disk(xpram_disks[i]);
+	}
+	unregister_blkdev(XPRAM_MAJOR, XPRAM_NAME);
+	platform_device_unregister(xpram_pdev);
+	platform_driver_unregister(&xpram_pdrv);
+}
+
+static int __init xpram_init(void)
+{
+	int rc;
+
+	/* Find out size of expanded memory. */
+	if (xpram_present() != 0) {
+		pr_err("No expanded memory available\n");
+		return -ENODEV;
+	}
+	xpram_pages = xpram_highest_page_index() + 1;
+	pr_info("  %u pages expanded memory found (%lu KB).\n",
+		xpram_pages, (unsigned long) xpram_pages*4);
+	rc = xpram_setup_sizes(xpram_pages);
+	if (rc)
+		return rc;
+	rc = platform_driver_register(&xpram_pdrv);
+	if (rc)
+		return rc;
+	xpram_pdev = platform_device_register_simple(XPRAM_NAME, -1, NULL, 0);
+	if (IS_ERR(xpram_pdev)) {
+		rc = PTR_ERR(xpram_pdev);
+		goto fail_platform_driver_unregister;
+	}
+	rc = xpram_setup_blkdev();
+	if (rc)
+		goto fail_platform_device_unregister;
+	return 0;
+
+fail_platform_device_unregister:
+	platform_device_unregister(xpram_pdev);
+fail_platform_driver_unregister:
+	platform_driver_unregister(&xpram_pdrv);
+	return rc;
+}
+
+module_init(xpram_init);
+module_exit(xpram_exit);

drivers/s390/char/Kconfig

@@ -0,0 +1,191 @@
+comment "S/390 character device drivers"
+	depends on S390
+
+config TN3270
+	def_tristate y
+	prompt "Support for locally attached 3270 terminals"
+	depends on CCW
+	help
+	  Include support for IBM 3270 terminals.
+
+config TN3270_TTY
+	def_tristate y
+	prompt "Support for tty input/output on 3270 terminals"
+	depends on TN3270 && TTY
+	help
+	  Include support for using an IBM 3270 terminal as a Linux tty.
+
+config TN3270_FS
+	def_tristate m
+	prompt "Support for fullscreen applications on 3270 terminals"
+	depends on TN3270
+	help
+	  Include support for fullscreen applications on an IBM 3270 terminal.
+
+config TN3270_CONSOLE
+	def_bool y
+	prompt "Support for console on 3270 terminal"
+	depends on TN3270=y && TN3270_TTY=y
+	help
+	  Include support for using an IBM 3270 terminal as a Linux system
+	  console.  Available only if 3270 support is compiled in statically.
+
+config TN3215
+	def_bool y
+	prompt "Support for 3215 line mode terminal"
+	depends on CCW && TTY
+	help
+	  Include support for IBM 3215 line-mode terminals.
+
+config TN3215_CONSOLE
+	def_bool y
+	prompt "Support for console on 3215 line mode terminal"
+	depends on TN3215
+	help
+	  Include support for using an IBM 3215 line-mode terminal as a
+	  Linux system console.
+
+config CCW_CONSOLE
+	def_bool y if TN3215_CONSOLE || TN3270_CONSOLE
+
+config SCLP_TTY
+	def_bool y
+	prompt "Support for SCLP line mode terminal"
+	depends on S390 && TTY
+	help
+	  Include support for IBM SCLP line-mode terminals.
+
+config SCLP_CONSOLE
+	def_bool y
+	prompt "Support for console on SCLP line mode terminal"
+	depends on SCLP_TTY
+	help
+	  Include support for using an IBM HWC line-mode terminal as the Linux
+	  system console.
+
+config SCLP_VT220_TTY
+	def_bool y
+	prompt "Support for SCLP VT220-compatible terminal"
+	depends on S390 && TTY
+	help
+	  Include support for an IBM SCLP VT220-compatible terminal.
+
+config SCLP_VT220_CONSOLE
+	def_bool y
+	prompt "Support for console on SCLP VT220-compatible terminal"
+	depends on SCLP_VT220_TTY
+	help
+	  Include support for using an IBM SCLP VT220-compatible terminal as a
+	  Linux system console.
+
+config SCLP_CPI
+	def_tristate m
+	prompt "Control-Program Identification"
+	depends on S390
+	help
+	  This option enables the hardware console interface for system
+	  identification. This is commonly used for workload management and
+	  gives you a nice name for the system on the service element.
+	  Please select this option as a module since built-in operation is
+	  completely untested.
+	  You should only select this option if you know what you are doing,
+	  need this feature and intend to run your kernel in LPAR.
+
+config SCLP_ASYNC
+	def_tristate m
+	prompt "Support for Call Home via Asynchronous SCLP Records"
+	depends on S390
+	help
+	  This option enables the call home function, which is able to inform
+	  the service element and connected organisations about a kernel panic.
+	  You should only select this option if you know what you are doing,
+	  want for inform other people about your kernel panics,
+	  need this feature and intend to run your kernel in LPAR.
+
+config HMC_DRV
+	def_tristate m
+	prompt "Support for file transfers from HMC drive CD/DVD-ROM"
+	depends on S390 && 64BIT
+	select CRC16
+	help
+	  This option enables support for file transfers from a Hardware
+	  Management Console (HMC) drive CD/DVD-ROM. It is available as a
+	  module, called 'hmcdrv', and also as kernel built-in. There is one
+	  optional parameter for this module: cachesize=N, which modifies the
+	  transfer cache size from it's default value 0.5MB to N bytes. If N
+	  is zero, then no caching is performed.
+
+config S390_TAPE
+	def_tristate m
+	prompt "S/390 tape device support"
+	depends on CCW
+	help
+	  Select this option if you want to access channel-attached tape
+	  devices on IBM S/390 or zSeries.
+	  If you select this option you will also want to select at
+	  least one of the tape interface options and one of the tape
+	  hardware options in order to access a tape device.
+	  This option is also available as a module. The module will be
+	  called tape390 and include all selected interfaces and
+	  hardware drivers.
+
+comment "S/390 tape hardware support"
+	depends on S390_TAPE
+
+config S390_TAPE_34XX
+	def_tristate m
+	prompt "Support for 3480/3490 tape hardware"
+	depends on S390_TAPE
+	help
+	  Select this option if you want to access IBM 3480/3490 magnetic
+	  tape subsystems and 100% compatibles.
+	  It is safe to say "Y" here.
+
+config S390_TAPE_3590
+	def_tristate m
+	prompt "Support for 3590 tape hardware"
+	depends on S390_TAPE
+	help
+	  Select this option if you want to access IBM 3590 magnetic
+	  tape subsystems and 100% compatibles.
+	  It is safe to say "Y" here.
+
+config VMLOGRDR
+	def_tristate m
+	prompt "Support for the z/VM recording system services (VM only)"
+	depends on IUCV
+	help
+	  Select this option if you want to be able to receive records collected
+	  by the z/VM recording system services, eg. from *LOGREC, *ACCOUNT or
+	  *SYMPTOM.
+	  This driver depends on the IUCV support driver.
+
+config VMCP
+	def_bool y
+	prompt "Support for the z/VM CP interface"
+	depends on S390
+	help
+	  Select this option if you want to be able to interact with the control
+	  program on z/VM
+
+config MONREADER
+	def_tristate m
+	prompt "API for reading z/VM monitor service records"
+	depends on IUCV
+	help
+	  Character device driver for reading z/VM monitor service records
+
+config MONWRITER
+	def_tristate m
+	prompt "API for writing z/VM monitor service records"
+	depends on S390
+	help
+	  Character device driver for writing z/VM monitor service records
+
+config S390_VMUR
+	def_tristate m
+	prompt "z/VM unit record device driver"
+	depends on S390
+	help
+	  Character device driver for z/VM reader, puncher and printer.
+

drivers/s390/char/Makefile

@@ -0,0 +1,38 @@
+#
+# S/390 character devices
+#
+
+obj-y += ctrlchar.o keyboard.o defkeymap.o sclp.o sclp_rw.o sclp_quiesce.o \
+	 sclp_cmd.o sclp_config.o sclp_cpi_sys.o sclp_ocf.o sclp_ctl.o \
+	 sclp_early.o
+
+obj-$(CONFIG_TN3270) += raw3270.o
+obj-$(CONFIG_TN3270_CONSOLE) += con3270.o
+obj-$(CONFIG_TN3270_TTY) += tty3270.o
+obj-$(CONFIG_TN3270_FS) += fs3270.o
+
+obj-$(CONFIG_TN3215) += con3215.o
+
+obj-$(CONFIG_SCLP_TTY) += sclp_tty.o
+obj-$(CONFIG_SCLP_CONSOLE) += sclp_con.o
+obj-$(CONFIG_SCLP_VT220_TTY) += sclp_vt220.o
+obj-$(CONFIG_SCLP_CPI) += sclp_cpi.o
+obj-$(CONFIG_SCLP_ASYNC) += sclp_async.o
+
+obj-$(CONFIG_VMLOGRDR) += vmlogrdr.o
+obj-$(CONFIG_VMCP) += vmcp.o
+
+tape-$(CONFIG_PROC_FS) += tape_proc.o
+tape-objs := tape_core.o tape_std.o tape_char.o $(tape-y)
+obj-$(CONFIG_S390_TAPE) += tape.o tape_class.o
+obj-$(CONFIG_S390_TAPE_34XX) += tape_34xx.o
+obj-$(CONFIG_S390_TAPE_3590) += tape_3590.o
+obj-$(CONFIG_MONREADER) += monreader.o
+obj-$(CONFIG_MONWRITER) += monwriter.o
+obj-$(CONFIG_S390_VMUR) += vmur.o
+
+zcore_mod-objs := sclp_sdias.o zcore.o
+obj-$(CONFIG_CRASH_DUMP) += zcore_mod.o
+
+hmcdrv-objs := hmcdrv_mod.o hmcdrv_dev.o hmcdrv_ftp.o hmcdrv_cache.o diag_ftp.o sclp_ftp.o
+obj-$(CONFIG_HMC_DRV) += hmcdrv.o

drivers/s390/char/con3270.c

@@ -0,0 +1,638 @@
+/*
+ * IBM/3270 Driver - console view.
+ *
+ * Author(s):
+ *   Original 3270 Code for 2.4 written by Richard Hitt (UTS Global)
+ *   Rewritten for 2.5 by Martin Schwidefsky <schwidefsky@de.ibm.com>
+ *     Copyright IBM Corp. 2003, 2009
+ */
+
+#include <linux/module.h>
+#include <linux/console.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/list.h>
+#include <linux/types.h>
+#include <linux/slab.h>
+#include <linux/err.h>
+#include <linux/reboot.h>
+
+#include <asm/ccwdev.h>
+#include <asm/cio.h>
+#include <asm/cpcmd.h>
+#include <asm/ebcdic.h>
+
+#include "raw3270.h"
+#include "tty3270.h"
+#include "ctrlchar.h"
+
+#define CON3270_OUTPUT_BUFFER_SIZE 1024
+#define CON3270_STRING_PAGES 4
+
+static struct raw3270_fn con3270_fn;
+
+static bool auto_update = 1;
+module_param(auto_update, bool, 0);
+
+/*
+ * Main 3270 console view data structure.
+ */
+struct con3270 {
+	struct raw3270_view view;
+	struct list_head freemem;	/* list of free memory for strings. */
+
+	/* Output stuff. */
+	struct list_head lines;		/* list of lines. */
+	struct list_head update;	/* list of lines to update. */
+	int line_nr;			/* line number for next update. */
+	int nr_lines;			/* # lines in list. */
+	int nr_up;			/* # lines up in history. */
+	unsigned long update_flags;	/* Update indication bits. */
+	struct string *cline;		/* current output line. */
+	struct string *status;		/* last line of display. */
+	struct raw3270_request *write;	/* single write request. */
+	struct timer_list timer;
+
+	/* Input stuff. */
+	struct string *input;		/* input string for read request. */
+	struct raw3270_request *read;	/* single read request. */
+	struct raw3270_request *kreset;	/* single keyboard reset request. */
+	struct tasklet_struct readlet;	/* tasklet to issue read request. */
+};
+
+static struct con3270 *condev;
+
+/* con3270->update_flags. See con3270_update for details. */
+#define CON_UPDATE_ERASE	1	/* Use EWRITEA instead of WRITE. */
+#define CON_UPDATE_LIST		2	/* Update lines in tty3270->update. */
+#define CON_UPDATE_STATUS	4	/* Update status line. */
+#define CON_UPDATE_ALL		8	/* Recreate screen. */
+
+static void con3270_update(struct con3270 *);
+
+/*
+ * Setup timeout for a device. On timeout trigger an update.
+ */
+static void con3270_set_timer(struct con3270 *cp, int expires)
+{
+	if (expires == 0)
+		del_timer(&cp->timer);
+	else
+		mod_timer(&cp->timer, jiffies + expires);
+}
+
+/*
+ * The status line is the last line of the screen. It shows the string
+ * "console view" in the lower left corner and "Running"/"More..."/"Holding"
+ * in the lower right corner of the screen.
+ */
+static void
+con3270_update_status(struct con3270 *cp)
+{
+	char *str;
+
+	str = (cp->nr_up != 0) ? "History" : "Running";
+	memcpy(cp->status->string + 24, str, 7);
+	codepage_convert(cp->view.ascebc, cp->status->string + 24, 7);
+	cp->update_flags |= CON_UPDATE_STATUS;
+}
+
+static void
+con3270_create_status(struct con3270 *cp)
+{
+	static const unsigned char blueprint[] =
+		{ TO_SBA, 0, 0, TO_SF,TF_LOG,TO_SA,TAT_COLOR, TAC_GREEN,
+		  'c','o','n','s','o','l','e',' ','v','i','e','w',
+		  TO_RA,0,0,0,'R','u','n','n','i','n','g',TO_SF,TF_LOG };
+
+	cp->status = alloc_string(&cp->freemem, sizeof(blueprint));
+	/* Copy blueprint to status line */
+	memcpy(cp->status->string, blueprint, sizeof(blueprint));
+	/* Set TO_RA addresses. */
+	raw3270_buffer_address(cp->view.dev, cp->status->string + 1,
+			       cp->view.cols * (cp->view.rows - 1));
+	raw3270_buffer_address(cp->view.dev, cp->status->string + 21,
+			       cp->view.cols * cp->view.rows - 8);
+	/* Convert strings to ebcdic. */
+	codepage_convert(cp->view.ascebc, cp->status->string + 8, 12);
+	codepage_convert(cp->view.ascebc, cp->status->string + 24, 7);
+}
+
+/*
+ * Set output offsets to 3270 datastream fragment of a console string.
+ */
+static void
+con3270_update_string(struct con3270 *cp, struct string *s, int nr)
+{
+	if (s->len >= cp->view.cols - 5)
+		return;
+	raw3270_buffer_address(cp->view.dev, s->string + s->len - 3,
+			       cp->view.cols * (nr + 1));
+}
+
+/*
+ * Rebuild update list to print all lines.
+ */
+static void
+con3270_rebuild_update(struct con3270 *cp)
+{
+	struct string *s, *n;
+	int nr;
+
+	/* 
+	 * Throw away update list and create a new one,
+	 * containing all lines that will fit on the screen.
+	 */
+	list_for_each_entry_safe(s, n, &cp->update, update)
+		list_del_init(&s->update);
+	nr = cp->view.rows - 2 + cp->nr_up;
+	list_for_each_entry_reverse(s, &cp->lines, list) {
+		if (nr < cp->view.rows - 1)
+			list_add(&s->update, &cp->update);
+		if (--nr < 0)
+			break;
+	}
+	cp->line_nr = 0;
+	cp->update_flags |= CON_UPDATE_LIST;
+}
+
+/*
+ * Alloc string for size bytes. Free strings from history if necessary.
+ */
+static struct string *
+con3270_alloc_string(struct con3270 *cp, size_t size)
+{
+	struct string *s, *n;
+
+	s = alloc_string(&cp->freemem, size);
+	if (s)
+		return s;
+	list_for_each_entry_safe(s, n, &cp->lines, list) {
+		list_del(&s->list);
+		if (!list_empty(&s->update))
+			list_del(&s->update);
+		cp->nr_lines--;
+		if (free_string(&cp->freemem, s) >= size)
+			break;
+	}
+	s = alloc_string(&cp->freemem, size);
+	BUG_ON(!s);
+	if (cp->nr_up != 0 && cp->nr_up + cp->view.rows > cp->nr_lines) {
+		cp->nr_up = cp->nr_lines - cp->view.rows + 1;
+		con3270_rebuild_update(cp);
+		con3270_update_status(cp);
+	}
+	return s;
+}
+
+/*
+ * Write completion callback.
+ */
+static void
+con3270_write_callback(struct raw3270_request *rq, void *data)
+{
+	raw3270_request_reset(rq);
+	xchg(&((struct con3270 *) rq->view)->write, rq);
+}
+
+/*
+ * Update console display.
+ */
+static void
+con3270_update(struct con3270 *cp)
+{
+	struct raw3270_request *wrq;
+	char wcc, prolog[6];
+	unsigned long flags;
+	unsigned long updated;
+	struct string *s, *n;
+	int rc;
+
+	if (!auto_update && !raw3270_view_active(&cp->view))
+		return;
+	if (cp->view.dev)
+		raw3270_activate_view(&cp->view);
+
+	wrq = xchg(&cp->write, 0);
+	if (!wrq) {
+		con3270_set_timer(cp, 1);
+		return;
+	}
+
+	spin_lock_irqsave(&cp->view.lock, flags);
+	updated = 0;
+	if (cp->update_flags & CON_UPDATE_ALL) {
+		con3270_rebuild_update(cp);
+		con3270_update_status(cp);
+		cp->update_flags = CON_UPDATE_ERASE | CON_UPDATE_LIST |
+			CON_UPDATE_STATUS;
+	}
+	if (cp->update_flags & CON_UPDATE_ERASE) {
+		/* Use erase write alternate to initialize display. */
+		raw3270_request_set_cmd(wrq, TC_EWRITEA);
+		updated |= CON_UPDATE_ERASE;
+	} else
+		raw3270_request_set_cmd(wrq, TC_WRITE);
+
+	wcc = TW_NONE;
+	raw3270_request_add_data(wrq, &wcc, 1);
+
+	/*
+	 * Update status line.
+	 */
+	if (cp->update_flags & CON_UPDATE_STATUS)
+		if (raw3270_request_add_data(wrq, cp->status->string,
+					     cp->status->len) == 0)
+			updated |= CON_UPDATE_STATUS;
+
+	if (cp->update_flags & CON_UPDATE_LIST) {
+		prolog[0] = TO_SBA;
+		prolog[3] = TO_SA;
+		prolog[4] = TAT_COLOR;
+		prolog[5] = TAC_TURQ;
+		raw3270_buffer_address(cp->view.dev, prolog + 1,
+				       cp->view.cols * cp->line_nr);
+		raw3270_request_add_data(wrq, prolog, 6);
+		/* Write strings in the update list to the screen. */
+		list_for_each_entry_safe(s, n, &cp->update, update) {
+			if (s != cp->cline)
+				con3270_update_string(cp, s, cp->line_nr);
+			if (raw3270_request_add_data(wrq, s->string,
+						     s->len) != 0)
+				break;
+			list_del_init(&s->update);
+			if (s != cp->cline)
+				cp->line_nr++;
+		}
+		if (list_empty(&cp->update))
+			updated |= CON_UPDATE_LIST;
+	}
+	wrq->callback = con3270_write_callback;
+	rc = raw3270_start(&cp->view, wrq);
+	if (rc == 0) {
+		cp->update_flags &= ~updated;
+		if (cp->update_flags)
+			con3270_set_timer(cp, 1);
+	} else {
+		raw3270_request_reset(wrq);
+		xchg(&cp->write, wrq);
+	}
+	spin_unlock_irqrestore(&cp->view.lock, flags);
+}
+
+/*
+ * Read tasklet.
+ */
+static void
+con3270_read_tasklet(struct raw3270_request *rrq)
+{
+	static char kreset_data = TW_KR;
+	struct con3270 *cp;
+	unsigned long flags;
+	int nr_up, deactivate;
+
+	cp = (struct con3270 *) rrq->view;
+	spin_lock_irqsave(&cp->view.lock, flags);
+	nr_up = cp->nr_up;
+	deactivate = 0;
+	/* Check aid byte. */
+	switch (cp->input->string[0]) {
+	case 0x7d:	/* enter: jump to bottom. */
+		nr_up = 0;
+		break;
+	case 0xf3:	/* PF3: deactivate the console view. */
+		deactivate = 1;
+		break;
+	case 0x6d:	/* clear: start from scratch. */
+		cp->update_flags = CON_UPDATE_ALL;
+		con3270_set_timer(cp, 1);
+		break;
+	case 0xf7:	/* PF7: do a page up in the console log. */
+		nr_up += cp->view.rows - 2;
+		if (nr_up + cp->view.rows - 1 > cp->nr_lines) {
+			nr_up = cp->nr_lines - cp->view.rows + 1;
+			if (nr_up < 0)
+				nr_up = 0;
+		}
+		break;
+	case 0xf8:	/* PF8: do a page down in the console log. */
+		nr_up -= cp->view.rows - 2;
+		if (nr_up < 0)
+			nr_up = 0;
+		break;
+	}
+	if (nr_up != cp->nr_up) {
+		cp->nr_up = nr_up;
+		con3270_rebuild_update(cp);
+		con3270_update_status(cp);
+		con3270_set_timer(cp, 1);
+	}
+	spin_unlock_irqrestore(&cp->view.lock, flags);
+
+	/* Start keyboard reset command. */
+	raw3270_request_reset(cp->kreset);
+	raw3270_request_set_cmd(cp->kreset, TC_WRITE);
+	raw3270_request_add_data(cp->kreset, &kreset_data, 1);
+	raw3270_start(&cp->view, cp->kreset);
+
+	if (deactivate)
+		raw3270_deactivate_view(&cp->view);
+
+	raw3270_request_reset(rrq);
+	xchg(&cp->read, rrq);
+	raw3270_put_view(&cp->view);
+}
+
+/*
+ * Read request completion callback.
+ */
+static void
+con3270_read_callback(struct raw3270_request *rq, void *data)
+{
+	raw3270_get_view(rq->view);
+	/* Schedule tasklet to pass input to tty. */
+	tasklet_schedule(&((struct con3270 *) rq->view)->readlet);
+}
+
+/*
+ * Issue a read request. Called only from interrupt function.
+ */
+static void
+con3270_issue_read(struct con3270 *cp)
+{
+	struct raw3270_request *rrq;
+	int rc;
+
+	rrq = xchg(&cp->read, 0);
+	if (!rrq)
+		/* Read already scheduled. */
+		return;
+	rrq->callback = con3270_read_callback;
+	rrq->callback_data = cp;
+	raw3270_request_set_cmd(rrq, TC_READMOD);
+	raw3270_request_set_data(rrq, cp->input->string, cp->input->len);
+	/* Issue the read modified request. */
+	rc = raw3270_start_irq(&cp->view, rrq);
+	if (rc)
+		raw3270_request_reset(rrq);
+}
+
+/*
+ * Switch to the console view.
+ */
+static int
+con3270_activate(struct raw3270_view *view)
+{
+	struct con3270 *cp;
+
+	cp = (struct con3270 *) view;
+	cp->update_flags = CON_UPDATE_ALL;
+	con3270_set_timer(cp, 1);
+	return 0;
+}
+
+static void
+con3270_deactivate(struct raw3270_view *view)
+{
+	struct con3270 *cp;
+
+	cp = (struct con3270 *) view;
+	del_timer(&cp->timer);
+}
+
+static int
+con3270_irq(struct con3270 *cp, struct raw3270_request *rq, struct irb *irb)
+{
+	/* Handle ATTN. Schedule tasklet to read aid. */
+	if (irb->scsw.cmd.dstat & DEV_STAT_ATTENTION)
+		con3270_issue_read(cp);
+
+	if (rq) {
+		if (irb->scsw.cmd.dstat & DEV_STAT_UNIT_CHECK)
+			rq->rc = -EIO;
+		else
+			/* Normal end. Copy residual count. */
+			rq->rescnt = irb->scsw.cmd.count;
+	}
+	return RAW3270_IO_DONE;
+}
+
+/* Console view to a 3270 device. */
+static struct raw3270_fn con3270_fn = {
+	.activate = con3270_activate,
+	.deactivate = con3270_deactivate,
+	.intv = (void *) con3270_irq
+};
+
+static inline void
+con3270_cline_add(struct con3270 *cp)
+{
+	if (!list_empty(&cp->cline->list))
+		/* Already added. */
+		return;
+	list_add_tail(&cp->cline->list, &cp->lines);
+	cp->nr_lines++;
+	con3270_rebuild_update(cp);
+}
+
+static inline void
+con3270_cline_insert(struct con3270 *cp, unsigned char c)
+{
+	cp->cline->string[cp->cline->len++] = 
+		cp->view.ascebc[(c < ' ') ? ' ' : c];
+	if (list_empty(&cp->cline->update)) {
+		list_add_tail(&cp->cline->update, &cp->update);
+		cp->update_flags |= CON_UPDATE_LIST;
+	}
+}
+
+static inline void
+con3270_cline_end(struct con3270 *cp)
+{
+	struct string *s;
+	unsigned int size;
+
+	/* Copy cline. */
+	size = (cp->cline->len < cp->view.cols - 5) ?
+		cp->cline->len + 4 : cp->view.cols;
+	s = con3270_alloc_string(cp, size);
+	memcpy(s->string, cp->cline->string, cp->cline->len);
+	if (s->len < cp->view.cols - 5) {
+		s->string[s->len - 4] = TO_RA;
+		s->string[s->len - 1] = 0;
+	} else {
+		while (--size > cp->cline->len)
+			s->string[size] = cp->view.ascebc[' '];
+	}
+	/* Replace cline with allocated line s and reset cline. */
+	list_add(&s->list, &cp->cline->list);
+	list_del_init(&cp->cline->list);
+	if (!list_empty(&cp->cline->update)) {
+		list_add(&s->update, &cp->cline->update);
+		list_del_init(&cp->cline->update);
+	}
+	cp->cline->len = 0;
+}
+
+/*
+ * Write a string to the 3270 console
+ */
+static void
+con3270_write(struct console *co, const char *str, unsigned int count)
+{
+	struct con3270 *cp;
+	unsigned long flags;
+	unsigned char c;
+
+	cp = condev;
+	spin_lock_irqsave(&cp->view.lock, flags);
+	while (count-- > 0) {
+		c = *str++;
+		if (cp->cline->len == 0)
+			con3270_cline_add(cp);
+		if (c != '\n')
+			con3270_cline_insert(cp, c);
+		if (c == '\n' || cp->cline->len >= cp->view.cols)
+			con3270_cline_end(cp);
+	}
+	/* Setup timer to output current console buffer after 1/10 second */
+	cp->nr_up = 0;
+	if (cp->view.dev && !timer_pending(&cp->timer))
+		con3270_set_timer(cp, HZ/10);
+	spin_unlock_irqrestore(&cp->view.lock,flags);
+}
+
+static struct tty_driver *
+con3270_device(struct console *c, int *index)
+{
+	*index = c->index;
+	return tty3270_driver;
+}
+
+/*
+ * Wait for end of write request.
+ */
+static void
+con3270_wait_write(struct con3270 *cp)
+{
+	while (!cp->write) {
+		raw3270_wait_cons_dev(cp->view.dev);
+		barrier();
+	}
+}
+
+/*
+ * panic() calls con3270_flush through a panic_notifier
+ * before the system enters a disabled, endless loop.
+ */
+static void
+con3270_flush(void)
+{
+	struct con3270 *cp;
+	unsigned long flags;
+
+	cp = condev;
+	if (!cp->view.dev)
+		return;
+	raw3270_pm_unfreeze(&cp->view);
+	raw3270_activate_view(&cp->view);
+	spin_lock_irqsave(&cp->view.lock, flags);
+	con3270_wait_write(cp);
+	cp->nr_up = 0;
+	con3270_rebuild_update(cp);
+	con3270_update_status(cp);
+	while (cp->update_flags != 0) {
+		spin_unlock_irqrestore(&cp->view.lock, flags);
+		con3270_update(cp);
+		spin_lock_irqsave(&cp->view.lock, flags);
+		con3270_wait_write(cp);
+	}
+	spin_unlock_irqrestore(&cp->view.lock, flags);
+}
+
+static int con3270_notify(struct notifier_block *self,
+			  unsigned long event, void *data)
+{
+	con3270_flush();
+	return NOTIFY_OK;
+}
+
+static struct notifier_block on_panic_nb = {
+	.notifier_call = con3270_notify,
+	.priority = 0,
+};
+
+static struct notifier_block on_reboot_nb = {
+	.notifier_call = con3270_notify,
+	.priority = 0,
+};
+
+/*
+ *  The console structure for the 3270 console
+ */
+static struct console con3270 = {
+	.name	 = "tty3270",
+	.write	 = con3270_write,
+	.device	 = con3270_device,
+	.flags	 = CON_PRINTBUFFER,
+};
+
+/*
+ * 3270 console initialization code called from console_init().
+ */
+static int __init
+con3270_init(void)
+{
+	struct raw3270 *rp;
+	void *cbuf;
+	int i;
+
+	/* Check if 3270 is to be the console */
+	if (!CONSOLE_IS_3270)
+		return -ENODEV;
+
+	/* Set the console mode for VM */
+	if (MACHINE_IS_VM) {
+		cpcmd("TERM CONMODE 3270", NULL, 0, NULL);
+		cpcmd("TERM AUTOCR OFF", NULL, 0, NULL);
+	}
+
+	rp = raw3270_setup_console();
+	if (IS_ERR(rp))
+		return PTR_ERR(rp);
+
+	condev = kzalloc(sizeof(struct con3270), GFP_KERNEL | GFP_DMA);
+	condev->view.dev = rp;
+
+	condev->read = raw3270_request_alloc(0);
+	condev->read->callback = con3270_read_callback;
+	condev->read->callback_data = condev;
+	condev->write = raw3270_request_alloc(CON3270_OUTPUT_BUFFER_SIZE);
+	condev->kreset = raw3270_request_alloc(1);
+
+	INIT_LIST_HEAD(&condev->lines);
+	INIT_LIST_HEAD(&condev->update);
+	setup_timer(&condev->timer, (void (*)(unsigned long)) con3270_update,
+		    (unsigned long) condev);
+	tasklet_init(&condev->readlet, 
+		     (void (*)(unsigned long)) con3270_read_tasklet,
+		     (unsigned long) condev->read);
+
+	raw3270_add_view(&condev->view, &con3270_fn, 1);
+
+	INIT_LIST_HEAD(&condev->freemem);
+	for (i = 0; i < CON3270_STRING_PAGES; i++) {
+		cbuf = (void *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+		add_string_memory(&condev->freemem, cbuf, PAGE_SIZE);
+	}
+	condev->cline = alloc_string(&condev->freemem, condev->view.cols);
+	condev->cline->len = 0;
+	con3270_create_status(condev);
+	condev->input = alloc_string(&condev->freemem, 80);
+	atomic_notifier_chain_register(&panic_notifier_list, &on_panic_nb);
+	register_reboot_notifier(&on_reboot_nb);
+	register_console(&con3270);
+	return 0;
+}
+
+console_initcall(con3270_init);

drivers/s390/char/ctrlchar.c

@@ -0,0 +1,72 @@
+/*
+ *  Unified handling of special chars.
+ *
+ *    Copyright IBM Corp. 2001
+ *    Author(s): Fritz Elfert <felfert@millenux.com> <elfert@de.ibm.com>
+ *
+ */
+
+#include <linux/stddef.h>
+#include <asm/errno.h>
+#include <linux/sysrq.h>
+#include <linux/ctype.h>
+
+#include "ctrlchar.h"
+
+#ifdef CONFIG_MAGIC_SYSRQ
+static int ctrlchar_sysrq_key;
+
+static void
+ctrlchar_handle_sysrq(struct work_struct *work)
+{
+	handle_sysrq(ctrlchar_sysrq_key);
+}
+
+static DECLARE_WORK(ctrlchar_work, ctrlchar_handle_sysrq);
+#endif
+
+
+/**
+ * Check for special chars at start of input.
+ *
+ * @param buf Console input buffer.
+ * @param len Length of valid data in buffer.
+ * @param tty The tty struct for this console.
+ * @return CTRLCHAR_NONE, if nothing matched,
+ *         CTRLCHAR_SYSRQ, if sysrq was encountered
+ *         otherwise char to be inserted logically or'ed
+ *         with CTRLCHAR_CTRL
+ */
+unsigned int
+ctrlchar_handle(const unsigned char *buf, int len, struct tty_struct *tty)
+{
+	if ((len < 2) || (len > 3))
+		return CTRLCHAR_NONE;
+
+	/* hat is 0xb1 in codepage 037 (US etc.) and thus */
+	/* converted to 0x5e in ascii ('^') */
+	if ((buf[0] != '^') && (buf[0] != '\252'))
+		return CTRLCHAR_NONE;
+
+#ifdef CONFIG_MAGIC_SYSRQ
+	/* racy */
+	if (len == 3 && buf[1] == '-') {
+		ctrlchar_sysrq_key = buf[2];
+		schedule_work(&ctrlchar_work);
+		return CTRLCHAR_SYSRQ;
+	}
+#endif
+
+	if (len != 2)
+		return CTRLCHAR_NONE;
+
+	switch (tolower(buf[1])) {
+	case 'c':
+		return INTR_CHAR(tty) | CTRLCHAR_CTRL;
+	case 'd':
+		return EOF_CHAR(tty)  | CTRLCHAR_CTRL;
+	case 'z':
+		return SUSP_CHAR(tty) | CTRLCHAR_CTRL;
+	}
+	return CTRLCHAR_NONE;
+}

drivers/s390/char/ctrlchar.h

@@ -0,0 +1,19 @@
+/*
+ *  Unified handling of special chars.
+ *
+ *    Copyright IBM Corp. 2001
+ *    Author(s): Fritz Elfert <felfert@millenux.com> <elfert@de.ibm.com>
+ *
+ */
+
+#include <linux/tty.h>
+
+extern unsigned int
+ctrlchar_handle(const unsigned char *buf, int len, struct tty_struct *tty);
+
+
+#define CTRLCHAR_NONE  (1 << 8)
+#define CTRLCHAR_CTRL  (2 << 8)
+#define CTRLCHAR_SYSRQ (3 << 8)
+
+#define CTRLCHAR_MASK (~0xffu)

drivers/s390/char/defkeymap.c

@@ -0,0 +1,158 @@
+
+/* Do not edit this file! It was automatically generated by   */
+/*    loadkeys --mktable defkeymap.map > defkeymap.c          */
+
+#include <linux/types.h>
+#include <linux/keyboard.h>
+#include <linux/kd.h>
+#include <linux/kbd_kern.h>
+#include <linux/kbd_diacr.h>
+
+u_short plain_map[NR_KEYS] = {
+	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,
+	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,
+	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,
+	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,
+	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,
+	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,
+	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,
+	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,	0xf000,
+	0xf020,	0xf000,	0xf0e2,	0xf0e4,	0xf0e0,	0xf0e1,	0xf0e3,	0xf0e5,
+	0xf0e7,	0xf0f1,	0xf0a2,	0xf02e,	0xf03c,	0xf028,	0xf02b,	0xf07c,
+	0xf026,	0xf0e9,	0xf0e2,	0xf0eb,	0xf0e8,	0xf0ed,	0xf0ee,	0xf0ef,
+	0xf0ec,	0xf0df,	0xf021,	0xf024,	0xf02a,	0xf029,	0xf03b,	0xf0ac,
+	0xf02d,	0xf02f,	0xf0c2,	0xf0c4,	0xf0c0,	0xf0c1,	0xf0c3,	0xf0c5,
+	0xf0c7,	0xf0d1,	0xf0a6,	0xf02c,	0xf025,	0xf05f,	0xf03e,	0xf03f,
+	0xf0f8,	0xf0c9,	0xf0ca,	0xf0cb,	0xf0c8,	0xf0cd,	0xf0ce,	0xf0cf,
+	0xf0cc,	0xf060,	0xf03a,	0xf023,	0xf040,	0xf027,	0xf03d,	0xf022,
+};
+
+static u_short shift_map[NR_KEYS] = {
+	0xf0d8,	0xf061,	0xf062,	0xf063,	0xf064,	0xf065,	0xf066,	0xf067,
+	0xf068,	0xf069,	0xf0ab,	0xf0bb,	0xf0f0,	0xf0fd,	0xf0fe,	0xf0b1,
+	0xf0b0,	0xf06a,	0xf06b,	0xf06c,	0xf06d,	0xf06e,	0xf06f,	0xf070,
+	0xf071,	0xf072,	0xf000,	0xf000,	0xf0e6,	0xf0b8,	0xf0c6,	0xf0a4,
+	0xf0b5,	0xf07e,	0xf073,	0xf074,	0xf075,	0xf076,	0xf077,	0xf078,
+	0xf079,	0xf07a,	0xf0a1,	0xf0bf,	0xf0d0,	0xf0dd,	0xf0de,	0xf0ae,
+	0xf402,	0xf0a3,	0xf0a5,	0xf0b7,	0xf0a9,	0xf0a7,	0xf0b6,	0xf0bc,
+	0xf0bd,	0xf0be,	0xf05b,	0xf05d,	0xf000,	0xf0a8,	0xf0b4,	0xf0d7,
+	0xf07b,	0xf041,	0xf042,	0xf043,	0xf044,	0xf045,	0xf046,	0xf047,
+	0xf048,	0xf049,	0xf000,	0xf0f4,	0xf0f6,	0xf0f2,	0xf0f3,	0xf0f5,
+	0xf07d,	0xf04a,	0xf04b,	0xf04c,	0xf04d,	0xf04e,	0xf04f,	0xf050,
+	0xf051,	0xf052,	0xf0b9,	0xf0fb,	0xf0fc,	0xf0f9,	0xf0fa,	0xf0ff,
+	0xf05c,	0xf0f7,	0xf053,	0xf054,	0xf055,	0xf056,	0xf057,	0xf058,
+	0xf059,	0xf05a,	0xf0b2,	0xf0d4,	0xf0d6,	0xf0d2,	0xf0d3,	0xf0d5,
+	0xf030,	0xf031,	0xf032,	0xf033,	0xf034,	0xf035,	0xf036,	0xf037,
+	0xf038,	0xf039,	0xf0b3,	0xf0db,	0xf0dc,	0xf0d9,	0xf0da,	0xf000,
+};
+
+static u_short ctrl_map[NR_KEYS] = {
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf11f,	0xf120,	0xf121,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf01a,	0xf003,	0xf212,	0xf004,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf109,	0xf10a,	0xf206,	0xf00a,	0xf200,	0xf200,
+};
+
+static u_short shift_ctrl_map[NR_KEYS] = {
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf10c,	0xf10d,	0xf10e,	0xf10f,	0xf110,	0xf111,	0xf112,
+	0xf113,	0xf11e,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+	0xf200,	0xf100,	0xf101,	0xf211,	0xf103,	0xf104,	0xf105,	0xf20b,
+	0xf20a,	0xf108,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,	0xf200,
+};
+
+ushort *key_maps[MAX_NR_KEYMAPS] = {
+	plain_map, shift_map, NULL, NULL,
+	ctrl_map, shift_ctrl_map, NULL,
+};
+
+unsigned int keymap_count = 4;
+
+
+/*
+ * Philosophy: most people do not define more strings, but they who do
+ * often want quite a lot of string space. So, we statically allocate
+ * the default and allocate dynamically in chunks of 512 bytes.
+ */
+
+char func_buf[] = {
+	'\033', '[', '[', 'A', 0, 
+	'\033', '[', '[', 'B', 0, 
+	'\033', '[', '[', 'C', 0, 
+	'\033', '[', '[', 'D', 0, 
+	'\033', '[', '[', 'E', 0, 
+	'\033', '[', '1', '7', '~', 0, 
+	'\033', '[', '1', '8', '~', 0, 
+	'\033', '[', '1', '9', '~', 0, 
+	'\033', '[', '2', '0', '~', 0, 
+	'\033', '[', '2', '1', '~', 0, 
+	'\033', '[', '2', '3', '~', 0, 
+	'\033', '[', '2', '4', '~', 0, 
+	'\033', '[', '2', '5', '~', 0, 
+	'\033', '[', '2', '6', '~', 0, 
+	'\033', '[', '2', '8', '~', 0, 
+	'\033', '[', '2', '9', '~', 0, 
+	'\033', '[', '3', '1', '~', 0, 
+	'\033', '[', '3', '2', '~', 0, 
+	'\033', '[', '3', '3', '~', 0, 
+	'\033', '[', '3', '4', '~', 0, 
+};
+
+
+char *funcbufptr = func_buf;
+int funcbufsize = sizeof(func_buf);
+int funcbufleft = 0;          /* space left */
+
+char *func_table[MAX_NR_FUNC] = {
+	func_buf + 0,
+	func_buf + 5,
+	func_buf + 10,
+	func_buf + 15,
+	func_buf + 20,
+	func_buf + 25,
+	func_buf + 31,
+	func_buf + 37,
+	func_buf + 43,
+	func_buf + 49,
+	func_buf + 55,
+	func_buf + 61,
+	func_buf + 67,
+	func_buf + 73,
+	func_buf + 79,
+	func_buf + 85,
+	func_buf + 91,
+	func_buf + 97,
+	func_buf + 103,
+	func_buf + 109,
+	NULL,
+};
+
+struct kbdiacruc accent_table[MAX_DIACR] = {
+	{'^', 'c', 0003},	{'^', 'd', 0004},
+	{'^', 'z', 0032},	{'^', 0012, 0000},
+};
+
+unsigned int accent_table_size = 4;

drivers/s390/char/defkeymap.map

@@ -0,0 +1,191 @@
+# Default keymap for 3270 (ebcdic codepage 037).
+keymaps 0-1,4-5
+
+keycode   0 = nul		Oslash
+keycode   1 = nul		a
+keycode   2 = nul		b
+keycode   3 = nul		c
+keycode   4 = nul		d
+keycode   5 = nul		e
+keycode   6 = nul		f
+keycode   7 = nul		g
+keycode   8 = nul		h
+keycode   9 = nul		i
+keycode  10 = nul		guillemotleft
+keycode  11 = nul		guillemotright
+keycode  12 = nul		eth
+keycode  13 = nul		yacute
+keycode  14 = nul		thorn
+keycode  15 = nul		plusminus
+keycode  16 = nul		degree
+keycode  17 = nul		j
+keycode  18 = nul		k
+keycode  19 = nul		l
+keycode  20 = nul		m
+keycode  21 = nul		n
+keycode  22 = nul		o
+keycode  23 = nul		p
+keycode  24 = nul		q
+keycode  25 = nul		r
+keycode  26 = nul		nul
+keycode  27 = nul		nul
+keycode  28 = nul		ae
+keycode  29 = nul		cedilla
+keycode  30 = nul		AE
+keycode  31 = nul		currency
+keycode  32 = nul		mu
+keycode  33 = nul		tilde
+keycode  34 = nul		s
+keycode  35 = nul		t
+keycode  36 = nul		u
+keycode  37 = nul		v
+keycode  38 = nul		w
+keycode  39 = nul		x
+keycode  40 = nul		y
+keycode  41 = nul		z
+keycode  42 = nul		exclamdown
+keycode  43 = nul		questiondown
+keycode  44 = nul		ETH
+keycode  45 = nul		Yacute
+keycode  46 = nul		THORN
+keycode  47 = nul		registered
+keycode  48 = nul		dead_circumflex
+keycode  49 = nul		sterling
+keycode  50 = nul		yen
+keycode  51 = nul		periodcentered
+keycode  52 = nul		copyright
+keycode  53 = nul		section
+keycode  54 = nul		paragraph
+keycode  55 = nul		onequarter
+keycode  56 = nul		onehalf
+keycode  57 = nul		threequarters
+keycode  58 = nul		bracketleft
+keycode  59 = nul		bracketright
+keycode  60 = nul		nul
+keycode  61 = nul		diaeresis
+keycode  62 = nul		acute
+keycode  63 = nul		multiply
+keycode  64 = space		braceleft
+keycode  65 = nul		A
+keycode  66 = acircumflex	B
+keycode  67 = adiaeresis	C
+keycode  68 = agrave		D
+keycode  69 = aacute		E
+keycode  70 = atilde		F
+keycode  71 = aring		G
+keycode  72 = ccedilla		H
+keycode  73 = ntilde		I
+keycode  74 = cent		nul
+keycode  75 = period		ocircumflex
+keycode  76 = less		odiaeresis
+keycode  77 = parenleft		ograve
+keycode  78 = plus		oacute
+keycode  79 = bar		otilde
+keycode  80 = ampersand		braceright
+keycode  81 = eacute		J
+keycode  82 = acircumflex	K
+keycode  83 = ediaeresis	L
+keycode  84 = egrave		M
+keycode  85 = iacute		N
+keycode  86 = icircumflex	O
+keycode  87 = idiaeresis	P
+keycode  88 = igrave		Q
+keycode  89 = ssharp		R
+keycode  90 = exclam		onesuperior
+keycode  91 = dollar		ucircumflex
+keycode  92 = asterisk		udiaeresis
+keycode  93 = parenright	ugrave
+keycode  94 = semicolon		uacute
+keycode  95 = notsign		ydiaeresis
+keycode  96 = minus		backslash
+keycode  97 = slash		division
+keycode  98 = Acircumflex	S
+keycode  99 = Adiaeresis	T
+keycode 100 = Agrave		U
+keycode 101 = Aacute		V
+keycode 102 = Atilde		W
+keycode 103 = Aring		X
+keycode 104 = Ccedilla		Y
+keycode 105 = Ntilde		Z
+keycode 106 = brokenbar		twosuperior
+keycode 107 = comma		Ocircumflex
+keycode 108 = percent		Odiaeresis
+keycode 109 = underscore	Ograve
+keycode 110 = greater		Oacute
+keycode 111 = question		Otilde
+keycode 112 = oslash		zero
+keycode 113 = Eacute		one
+keycode 114 = Ecircumflex	two
+keycode 115 = Ediaeresis	three
+keycode 116 = Egrave		four
+keycode 117 = Iacute		five
+keycode 118 = Icircumflex	six
+keycode 119 = Idiaeresis	seven
+keycode 120 = Igrave		eight
+keycode 121 = grave		nine
+keycode 122 = colon		threesuperior
+keycode 123 = numbersign	Ucircumflex
+keycode 124 = at		Udiaeresis
+keycode 125 = apostrophe	Ugrave
+keycode 126 = equal		Uacute
+keycode 127 = quotedbl		nul
+
+# AID keys
+control keycode  74 = F22
+control keycode  75 = F23
+control keycode  76 = F24
+control keycode 107 = Control_z		# PA3
+control keycode 108 = Control_c		# PA1
+control keycode 109 = KeyboardSignal	# Clear
+control keycode 110 = Control_d		# PA2
+control keycode 122 = F10
+control keycode 123 = F11		# F11
+control keycode 124 = Last_Console	# F12
+control keycode 125 = Linefeed
+shift control keycode  65 = F13
+shift control keycode  66 = F14
+shift control keycode  67 = F15
+shift control keycode  68 = F16
+shift control keycode  69 = F17
+shift control keycode  70 = F18
+shift control keycode  71 = F19
+shift control keycode  72 = F20
+shift control keycode  73 = F21
+shift control keycode 113 = F1
+shift control keycode 114 = F2
+shift control keycode 115 = Incr_Console
+shift control keycode 116 = F4
+shift control keycode 117 = F5
+shift control keycode 118 = F6
+shift control keycode 119 = Scroll_Backward
+shift control keycode 120 = Scroll_Forward
+shift control keycode 121 = F9
+
+string F1 = "\033[[A"
+string F2 = "\033[[B"
+string F3 = "\033[[C"
+string F4 = "\033[[D"
+string F5 = "\033[[E"
+string F6 = "\033[17~"
+string F7 = "\033[18~"
+string F8 = "\033[19~"
+string F9 = "\033[20~"
+string F10 = "\033[21~"
+string F11 = "\033[23~"
+string F12 = "\033[24~"
+string F13 = "\033[25~"
+string F14 = "\033[26~"
+string F15 = "\033[28~"
+string F16 = "\033[29~"
+string F17 = "\033[31~"
+string F18 = "\033[32~"
+string F19 = "\033[33~"
+string F20 = "\033[34~"
+# string F21 ??
+# string F22 ??
+# string F23 ??
+# string F24 ??
+compose '^' 'c' to Control_c
+compose '^' 'd' to Control_d
+compose '^' 'z' to Control_z
+compose '^' '\012' to nul

drivers/s390/char/diag_ftp.c

@@ -0,0 +1,237 @@
+/*
+ *    DIAGNOSE X'2C4' instruction based HMC FTP services, useable on z/VM
+ *
+ *    Copyright IBM Corp. 2013
+ *    Author(s): Ralf Hoppe (rhoppe@de.ibm.com)
+ *
+ */
+
+#define KMSG_COMPONENT "hmcdrv"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/irq.h>
+#include <linux/wait.h>
+#include <linux/string.h>
+#include <asm/ctl_reg.h>
+
+#include "hmcdrv_ftp.h"
+#include "diag_ftp.h"
+
+/* DIAGNOSE X'2C4' return codes in Ry */
+#define DIAG_FTP_RET_OK	0 /* HMC FTP started successfully */
+#define DIAG_FTP_RET_EBUSY	4 /* HMC FTP service currently busy */
+#define DIAG_FTP_RET_EIO	8 /* HMC FTP service I/O error */
+/* and an artificial extension */
+#define DIAG_FTP_RET_EPERM	2 /* HMC FTP service privilege error */
+
+/* FTP service status codes (after INTR at guest real location 133) */
+#define DIAG_FTP_STAT_OK	0U /* request completed successfully */
+#define DIAG_FTP_STAT_PGCC	4U /* program check condition */
+#define DIAG_FTP_STAT_PGIOE	8U /* paging I/O error */
+#define DIAG_FTP_STAT_TIMEOUT	12U /* timeout */
+#define DIAG_FTP_STAT_EBASE	16U /* base of error codes from SCLP */
+#define DIAG_FTP_STAT_LDFAIL	(DIAG_FTP_STAT_EBASE + 1U) /* failed */
+#define DIAG_FTP_STAT_LDNPERM	(DIAG_FTP_STAT_EBASE + 2U) /* not allowed */
+#define DIAG_FTP_STAT_LDRUNS	(DIAG_FTP_STAT_EBASE + 3U) /* runs */
+#define DIAG_FTP_STAT_LDNRUNS	(DIAG_FTP_STAT_EBASE + 4U) /* not runs */
+
+/**
+ * struct diag_ftp_ldfpl - load file FTP parameter list (LDFPL)
+ * @bufaddr: real buffer address (at 4k boundary)
+ * @buflen: length of buffer
+ * @offset: dir/file offset
+ * @intparm: interruption parameter (unused)
+ * @transferred: bytes transferred
+ * @fsize: file size, filled on GET
+ * @failaddr: failing address
+ * @spare: padding
+ * @fident: file name - ASCII
+ */
+struct diag_ftp_ldfpl {
+	u64 bufaddr;
+	u64 buflen;
+	u64 offset;
+	u64 intparm;
+	u64 transferred;
+	u64 fsize;
+	u64 failaddr;
+	u64 spare;
+	u8 fident[HMCDRV_FTP_FIDENT_MAX];
+} __packed;
+
+static DECLARE_COMPLETION(diag_ftp_rx_complete);
+static int diag_ftp_subcode;
+
+/**
+ * diag_ftp_handler() - FTP services IRQ handler
+ * @extirq: external interrupt (sub-) code
+ * @param32: 32-bit interruption parameter from &struct diag_ftp_ldfpl
+ * @param64: unused (for 64-bit interrupt parameters)
+ */
+static void diag_ftp_handler(struct ext_code extirq,
+			     unsigned int param32,
+			     unsigned long param64)
+{
+	if ((extirq.subcode >> 8) != 8)
+		return; /* not a FTP services sub-code */
+
+	inc_irq_stat(IRQEXT_FTP);
+	diag_ftp_subcode = extirq.subcode & 0xffU;
+	complete(&diag_ftp_rx_complete);
+}
+
+/**
+ * diag_ftp_2c4() - DIAGNOSE X'2C4' service call
+ * @fpl: pointer to prepared LDFPL
+ * @cmd: FTP command to be executed
+ *
+ * Performs a DIAGNOSE X'2C4' call with (input/output) FTP parameter list
+ * @fpl and FTP function code @cmd. In case of an error the function does
+ * nothing and returns an (negative) error code.
+ *
+ * Notes:
+ * 1. This function only initiates a transfer, so the caller must wait
+ *    for completion (asynchronous execution).
+ * 2. The FTP parameter list @fpl must be aligned to a double-word boundary.
+ * 3. fpl->bufaddr must be a real address, 4k aligned
+ */
+static int diag_ftp_2c4(struct diag_ftp_ldfpl *fpl,
+			enum hmcdrv_ftp_cmdid cmd)
+{
+	int rc;
+
+	asm volatile(
+		"	diag	%[addr],%[cmd],0x2c4\n"
+		"0:	j	2f\n"
+		"1:	la	%[rc],%[err]\n"
+		"2:\n"
+		EX_TABLE(0b, 1b)
+		: [rc] "=d" (rc), "+m" (*fpl)
+		: [cmd] "0" (cmd), [addr] "d" (virt_to_phys(fpl)),
+		  [err] "i" (DIAG_FTP_RET_EPERM)
+		: "cc");
+
+	switch (rc) {
+	case DIAG_FTP_RET_OK:
+		return 0;
+	case DIAG_FTP_RET_EBUSY:
+		return -EBUSY;
+	case DIAG_FTP_RET_EPERM:
+		return -EPERM;
+	case DIAG_FTP_RET_EIO:
+	default:
+		return -EIO;
+	}
+}
+
+/**
+ * diag_ftp_cmd() - executes a DIAG X'2C4' FTP command, targeting a HMC
+ * @ftp: pointer to FTP command specification
+ * @fsize: return of file size (or NULL if undesirable)
+ *
+ * Attention: Notice that this function is not reentrant - so the caller
+ * must ensure locking.
+ *
+ * Return: number of bytes read/written or a (negative) error code
+ */
+ssize_t diag_ftp_cmd(const struct hmcdrv_ftp_cmdspec *ftp, size_t *fsize)
+{
+	struct diag_ftp_ldfpl *ldfpl;
+	ssize_t len;
+#ifdef DEBUG
+	unsigned long start_jiffies;
+
+	pr_debug("starting DIAG X'2C4' on '%s', requesting %zd bytes\n",
+		 ftp->fname, ftp->len);
+	start_jiffies = jiffies;
+#endif
+	init_completion(&diag_ftp_rx_complete);
+
+	ldfpl = (void *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+	if (!ldfpl) {
+		len = -ENOMEM;
+		goto out;
+	}
+
+	len = strlcpy(ldfpl->fident, ftp->fname, sizeof(ldfpl->fident));
+	if (len >= HMCDRV_FTP_FIDENT_MAX) {
+		len = -EINVAL;
+		goto out_free;
+	}
+
+	ldfpl->transferred = 0;
+	ldfpl->fsize = 0;
+	ldfpl->offset = ftp->ofs;
+	ldfpl->buflen = ftp->len;
+	ldfpl->bufaddr = virt_to_phys(ftp->buf);
+
+	len = diag_ftp_2c4(ldfpl, ftp->id);
+	if (len)
+		goto out_free;
+
+	/*
+	 * There is no way to cancel the running diag X'2C4', the code
+	 * needs to wait unconditionally until the transfer is complete.
+	 */
+	wait_for_completion(&diag_ftp_rx_complete);
+
+#ifdef DEBUG
+	pr_debug("completed DIAG X'2C4' after %lu ms\n",
+		 (jiffies - start_jiffies) * 1000 / HZ);
+	pr_debug("status of DIAG X'2C4' is %u, with %lld/%lld bytes\n",
+		 diag_ftp_subcode, ldfpl->transferred, ldfpl->fsize);
+#endif
+
+	switch (diag_ftp_subcode) {
+	case DIAG_FTP_STAT_OK: /* success */
+		len = ldfpl->transferred;
+		if (fsize)
+			*fsize = ldfpl->fsize;
+		break;
+	case DIAG_FTP_STAT_LDNPERM:
+		len = -EPERM;
+		break;
+	case DIAG_FTP_STAT_LDRUNS:
+		len = -EBUSY;
+		break;
+	case DIAG_FTP_STAT_LDFAIL:
+		len = -ENOENT; /* no such file or media */
+		break;
+	default:
+		len = -EIO;
+		break;
+	}
+
+out_free:
+	free_page((unsigned long) ldfpl);
+out:
+	return len;
+}
+
+/**
+ * diag_ftp_startup() - startup of FTP services, when running on z/VM
+ *
+ * Return: 0 on success, else an (negative) error code
+ */
+int diag_ftp_startup(void)
+{
+	int rc;
+
+	rc = register_external_irq(EXT_IRQ_CP_SERVICE, diag_ftp_handler);
+	if (rc)
+		return rc;
+
+	ctl_set_bit(0, 63 - 22);
+	return 0;
+}
+
+/**
+ * diag_ftp_shutdown() - shutdown of FTP services, when running on z/VM
+ */
+void diag_ftp_shutdown(void)
+{
+	ctl_clear_bit(0, 63 - 22);
+	unregister_external_irq(EXT_IRQ_CP_SERVICE, diag_ftp_handler);
+}

drivers/s390/char/diag_ftp.h

@@ -0,0 +1,21 @@
+/*
+ *    DIAGNOSE X'2C4' instruction based SE/HMC FTP Services, useable on z/VM
+ *
+ *    Notice that all functions exported here are not reentrant.
+ *    So usage should be exclusive, ensured by the caller (e.g. using a
+ *    mutex).
+ *
+ *    Copyright IBM Corp. 2013
+ *    Author(s): Ralf Hoppe (rhoppe@de.ibm.com)
+ */
+
+#ifndef __DIAG_FTP_H__
+#define __DIAG_FTP_H__
+
+#include "hmcdrv_ftp.h"
+
+int diag_ftp_startup(void);
+void diag_ftp_shutdown(void);
+ssize_t diag_ftp_cmd(const struct hmcdrv_ftp_cmdspec *ftp, size_t *fsize);
+
+#endif	 /* __DIAG_FTP_H__ */

drivers/s390/char/fs3270.c

@@ -0,0 +1,575 @@
+/*
+ * IBM/3270 Driver - fullscreen driver.
+ *
+ * Author(s):
+ *   Original 3270 Code for 2.4 written by Richard Hitt (UTS Global)
+ *   Rewritten for 2.5/2.6 by Martin Schwidefsky <schwidefsky@de.ibm.com>
+ *     Copyright IBM Corp. 2003, 2009
+ */
+
+#include <linux/bootmem.h>
+#include <linux/console.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/compat.h>
+#include <linux/module.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/types.h>
+
+#include <asm/compat.h>
+#include <asm/ccwdev.h>
+#include <asm/cio.h>
+#include <asm/ebcdic.h>
+#include <asm/idals.h>
+
+#include "raw3270.h"
+#include "ctrlchar.h"
+
+static struct raw3270_fn fs3270_fn;
+
+struct fs3270 {
+	struct raw3270_view view;
+	struct pid *fs_pid;		/* Pid of controlling program. */
+	int read_command;		/* ccw command to use for reads. */
+	int write_command;		/* ccw command to use for writes. */
+	int attention;			/* Got attention. */
+	int active;			/* Fullscreen view is active. */
+	struct raw3270_request *init;	/* single init request. */
+	wait_queue_head_t wait;		/* Init & attention wait queue. */
+	struct idal_buffer *rdbuf;	/* full-screen-deactivate buffer */
+	size_t rdbuf_size;		/* size of data returned by RDBUF */
+};
+
+static DEFINE_MUTEX(fs3270_mutex);
+
+static void
+fs3270_wake_up(struct raw3270_request *rq, void *data)
+{
+	wake_up((wait_queue_head_t *) data);
+}
+
+static inline int
+fs3270_working(struct fs3270 *fp)
+{
+	/*
+	 * The fullscreen view is in working order if the view
+	 * has been activated AND the initial request is finished.
+	 */
+	return fp->active && raw3270_request_final(fp->init);
+}
+
+static int
+fs3270_do_io(struct raw3270_view *view, struct raw3270_request *rq)
+{
+	struct fs3270 *fp;
+	int rc;
+
+	fp = (struct fs3270 *) view;
+	rq->callback = fs3270_wake_up;
+	rq->callback_data = &fp->wait;
+
+	do {
+		if (!fs3270_working(fp)) {
+			/* Fullscreen view isn't ready yet. */
+			rc = wait_event_interruptible(fp->wait,
+						      fs3270_working(fp));
+			if (rc != 0)
+				break;
+		}
+		rc = raw3270_start(view, rq);
+		if (rc == 0) {
+			/* Started successfully. Now wait for completion. */
+			wait_event(fp->wait, raw3270_request_final(rq));
+		}
+	} while (rc == -EACCES);
+	return rc;
+}
+
+/*
+ * Switch to the fullscreen view.
+ */
+static void
+fs3270_reset_callback(struct raw3270_request *rq, void *data)
+{
+	struct fs3270 *fp;
+
+	fp = (struct fs3270 *) rq->view;
+	raw3270_request_reset(rq);
+	wake_up(&fp->wait);
+}
+
+static void
+fs3270_restore_callback(struct raw3270_request *rq, void *data)
+{
+	struct fs3270 *fp;
+
+	fp = (struct fs3270 *) rq->view;
+	if (rq->rc != 0 || rq->rescnt != 0) {
+		if (fp->fs_pid)
+			kill_pid(fp->fs_pid, SIGHUP, 1);
+	}
+	fp->rdbuf_size = 0;
+	raw3270_request_reset(rq);
+	wake_up(&fp->wait);
+}
+
+static int
+fs3270_activate(struct raw3270_view *view)
+{
+	struct fs3270 *fp;
+	char *cp;
+	int rc;
+
+	fp = (struct fs3270 *) view;
+
+	/* If an old init command is still running just return. */
+	if (!raw3270_request_final(fp->init))
+		return 0;
+
+	if (fp->rdbuf_size == 0) {
+		/* No saved buffer. Just clear the screen. */
+		raw3270_request_set_cmd(fp->init, TC_EWRITEA);
+		fp->init->callback = fs3270_reset_callback;
+	} else {
+		/* Restore fullscreen buffer saved by fs3270_deactivate. */
+		raw3270_request_set_cmd(fp->init, TC_EWRITEA);
+		raw3270_request_set_idal(fp->init, fp->rdbuf);
+		fp->init->ccw.count = fp->rdbuf_size;
+		cp = fp->rdbuf->data[0];
+		cp[0] = TW_KR;
+		cp[1] = TO_SBA;
+		cp[2] = cp[6];
+		cp[3] = cp[7];
+		cp[4] = TO_IC;
+		cp[5] = TO_SBA;
+		cp[6] = 0x40;
+		cp[7] = 0x40;
+		fp->init->rescnt = 0;
+		fp->init->callback = fs3270_restore_callback;
+	}
+	rc = fp->init->rc = raw3270_start_locked(view, fp->init);
+	if (rc)
+		fp->init->callback(fp->init, NULL);
+	else
+		fp->active = 1;
+	return rc;
+}
+
+/*
+ * Shutdown fullscreen view.
+ */
+static void
+fs3270_save_callback(struct raw3270_request *rq, void *data)
+{
+	struct fs3270 *fp;
+
+	fp = (struct fs3270 *) rq->view;
+
+	/* Correct idal buffer element 0 address. */
+	fp->rdbuf->data[0] -= 5;
+	fp->rdbuf->size += 5;
+
+	/*
+	 * If the rdbuf command failed or the idal buffer is
+	 * to small for the amount of data returned by the
+	 * rdbuf command, then we have no choice but to send
+	 * a SIGHUP to the application.
+	 */
+	if (rq->rc != 0 || rq->rescnt == 0) {
+		if (fp->fs_pid)
+			kill_pid(fp->fs_pid, SIGHUP, 1);
+		fp->rdbuf_size = 0;
+	} else
+		fp->rdbuf_size = fp->rdbuf->size - rq->rescnt;
+	raw3270_request_reset(rq);
+	wake_up(&fp->wait);
+}
+
+static void
+fs3270_deactivate(struct raw3270_view *view)
+{
+	struct fs3270 *fp;
+
+	fp = (struct fs3270 *) view;
+	fp->active = 0;
+
+	/* If an old init command is still running just return. */
+	if (!raw3270_request_final(fp->init))
+		return;
+
+	/* Prepare read-buffer request. */
+	raw3270_request_set_cmd(fp->init, TC_RDBUF);
+	/*
+	 * Hackish: skip first 5 bytes of the idal buffer to make
+	 * room for the TW_KR/TO_SBA/<address>/<address>/TO_IC sequence
+	 * in the activation command.
+	 */
+	fp->rdbuf->data[0] += 5;
+	fp->rdbuf->size -= 5;
+	raw3270_request_set_idal(fp->init, fp->rdbuf);
+	fp->init->rescnt = 0;
+	fp->init->callback = fs3270_save_callback;
+
+	/* Start I/O to read in the 3270 buffer. */
+	fp->init->rc = raw3270_start_locked(view, fp->init);
+	if (fp->init->rc)
+		fp->init->callback(fp->init, NULL);
+}
+
+static int
+fs3270_irq(struct fs3270 *fp, struct raw3270_request *rq, struct irb *irb)
+{
+	/* Handle ATTN. Set indication and wake waiters for attention. */
+	if (irb->scsw.cmd.dstat & DEV_STAT_ATTENTION) {
+		fp->attention = 1;
+		wake_up(&fp->wait);
+	}
+
+	if (rq) {
+		if (irb->scsw.cmd.dstat & DEV_STAT_UNIT_CHECK)
+			rq->rc = -EIO;
+		else
+			/* Normal end. Copy residual count. */
+			rq->rescnt = irb->scsw.cmd.count;
+	}
+	return RAW3270_IO_DONE;
+}
+
+/*
+ * Process reads from fullscreen 3270.
+ */
+static ssize_t
+fs3270_read(struct file *filp, char __user *data, size_t count, loff_t *off)
+{
+	struct fs3270 *fp;
+	struct raw3270_request *rq;
+	struct idal_buffer *ib;
+	ssize_t rc;
+	
+	if (count == 0 || count > 65535)
+		return -EINVAL;
+	fp = filp->private_data;
+	if (!fp)
+		return -ENODEV;
+	ib = idal_buffer_alloc(count, 0);
+	if (IS_ERR(ib))
+		return -ENOMEM;
+	rq = raw3270_request_alloc(0);
+	if (!IS_ERR(rq)) {
+		if (fp->read_command == 0 && fp->write_command != 0)
+			fp->read_command = 6;
+		raw3270_request_set_cmd(rq, fp->read_command ? : 2);
+		raw3270_request_set_idal(rq, ib);
+		rc = wait_event_interruptible(fp->wait, fp->attention);
+		fp->attention = 0;
+		if (rc == 0) {
+			rc = fs3270_do_io(&fp->view, rq);
+			if (rc == 0) {
+				count -= rq->rescnt;
+				if (idal_buffer_to_user(ib, data, count) != 0)
+					rc = -EFAULT;
+				else
+					rc = count;
+
+			}
+		}
+		raw3270_request_free(rq);
+	} else
+		rc = PTR_ERR(rq);
+	idal_buffer_free(ib);
+	return rc;
+}
+
+/*
+ * Process writes to fullscreen 3270.
+ */
+static ssize_t
+fs3270_write(struct file *filp, const char __user *data, size_t count, loff_t *off)
+{
+	struct fs3270 *fp;
+	struct raw3270_request *rq;
+	struct idal_buffer *ib;
+	int write_command;
+	ssize_t rc;
+
+	fp = filp->private_data;
+	if (!fp)
+		return -ENODEV;
+	ib = idal_buffer_alloc(count, 0);
+	if (IS_ERR(ib))
+		return -ENOMEM;
+	rq = raw3270_request_alloc(0);
+	if (!IS_ERR(rq)) {
+		if (idal_buffer_from_user(ib, data, count) == 0) {
+			write_command = fp->write_command ? : 1;
+			if (write_command == 5)
+				write_command = 13;
+			raw3270_request_set_cmd(rq, write_command);
+			raw3270_request_set_idal(rq, ib);
+			rc = fs3270_do_io(&fp->view, rq);
+			if (rc == 0)
+				rc = count - rq->rescnt;
+		} else
+			rc = -EFAULT;
+		raw3270_request_free(rq);
+	} else
+		rc = PTR_ERR(rq);
+	idal_buffer_free(ib);
+	return rc;
+}
+
+/*
+ * process ioctl commands for the tube driver
+ */
+static long
+fs3270_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
+{
+	char __user *argp;
+	struct fs3270 *fp;
+	struct raw3270_iocb iocb;
+	int rc;
+
+	fp = filp->private_data;
+	if (!fp)
+		return -ENODEV;
+	if (is_compat_task())
+		argp = compat_ptr(arg);
+	else
+		argp = (char __user *)arg;
+	rc = 0;
+	mutex_lock(&fs3270_mutex);
+	switch (cmd) {
+	case TUBICMD:
+		fp->read_command = arg;
+		break;
+	case TUBOCMD:
+		fp->write_command = arg;
+		break;
+	case TUBGETI:
+		rc = put_user(fp->read_command, argp);
+		break;
+	case TUBGETO:
+		rc = put_user(fp->write_command, argp);
+		break;
+	case TUBGETMOD:
+		iocb.model = fp->view.model;
+		iocb.line_cnt = fp->view.rows;
+		iocb.col_cnt = fp->view.cols;
+		iocb.pf_cnt = 24;
+		iocb.re_cnt = 20;
+		iocb.map = 0;
+		if (copy_to_user(argp, &iocb, sizeof(struct raw3270_iocb)))
+			rc = -EFAULT;
+		break;
+	}
+	mutex_unlock(&fs3270_mutex);
+	return rc;
+}
+
+/*
+ * Allocate fs3270 structure.
+ */
+static struct fs3270 *
+fs3270_alloc_view(void)
+{
+	struct fs3270 *fp;
+
+	fp = kzalloc(sizeof(struct fs3270),GFP_KERNEL);
+	if (!fp)
+		return ERR_PTR(-ENOMEM);
+	fp->init = raw3270_request_alloc(0);
+	if (IS_ERR(fp->init)) {
+		kfree(fp);
+		return ERR_PTR(-ENOMEM);
+	}
+	return fp;
+}
+
+/*
+ * Free fs3270 structure.
+ */
+static void
+fs3270_free_view(struct raw3270_view *view)
+{
+	struct fs3270 *fp;
+
+	fp = (struct fs3270 *) view;
+	if (fp->rdbuf)
+		idal_buffer_free(fp->rdbuf);
+	raw3270_request_free(((struct fs3270 *) view)->init);
+	kfree(view);
+}
+
+/*
+ * Unlink fs3270 data structure from filp.
+ */
+static void
+fs3270_release(struct raw3270_view *view)
+{
+	struct fs3270 *fp;
+
+	fp = (struct fs3270 *) view;
+	if (fp->fs_pid)
+		kill_pid(fp->fs_pid, SIGHUP, 1);
+}
+
+/* View to a 3270 device. Can be console, tty or fullscreen. */
+static struct raw3270_fn fs3270_fn = {
+	.activate = fs3270_activate,
+	.deactivate = fs3270_deactivate,
+	.intv = (void *) fs3270_irq,
+	.release = fs3270_release,
+	.free = fs3270_free_view
+};
+
+/*
+ * This routine is called whenever a 3270 fullscreen device is opened.
+ */
+static int
+fs3270_open(struct inode *inode, struct file *filp)
+{
+	struct fs3270 *fp;
+	struct idal_buffer *ib;
+	int minor, rc = 0;
+
+	if (imajor(file_inode(filp)) != IBM_FS3270_MAJOR)
+		return -ENODEV;
+	minor = iminor(file_inode(filp));
+	/* Check for minor 0 multiplexer. */
+	if (minor == 0) {
+		struct tty_struct *tty = get_current_tty();
+		if (!tty || tty->driver->major != IBM_TTY3270_MAJOR) {
+			tty_kref_put(tty);
+			return -ENODEV;
+		}
+		minor = tty->index;
+		tty_kref_put(tty);
+	}
+	mutex_lock(&fs3270_mutex);
+	/* Check if some other program is already using fullscreen mode. */
+	fp = (struct fs3270 *) raw3270_find_view(&fs3270_fn, minor);
+	if (!IS_ERR(fp)) {
+		raw3270_put_view(&fp->view);
+		rc = -EBUSY;
+		goto out;
+	}
+	/* Allocate fullscreen view structure. */
+	fp = fs3270_alloc_view();
+	if (IS_ERR(fp)) {
+		rc = PTR_ERR(fp);
+		goto out;
+	}
+
+	init_waitqueue_head(&fp->wait);
+	fp->fs_pid = get_pid(task_pid(current));
+	rc = raw3270_add_view(&fp->view, &fs3270_fn, minor);
+	if (rc) {
+		fs3270_free_view(&fp->view);
+		goto out;
+	}
+
+	/* Allocate idal-buffer. */
+	ib = idal_buffer_alloc(2*fp->view.rows*fp->view.cols + 5, 0);
+	if (IS_ERR(ib)) {
+		raw3270_put_view(&fp->view);
+		raw3270_del_view(&fp->view);
+		rc = PTR_ERR(ib);
+		goto out;
+	}
+	fp->rdbuf = ib;
+
+	rc = raw3270_activate_view(&fp->view);
+	if (rc) {
+		raw3270_put_view(&fp->view);
+		raw3270_del_view(&fp->view);
+		goto out;
+	}
+	nonseekable_open(inode, filp);
+	filp->private_data = fp;
+out:
+	mutex_unlock(&fs3270_mutex);
+	return rc;
+}
+
+/*
+ * This routine is called when the 3270 tty is closed. We wait
+ * for the remaining request to be completed. Then we clean up.
+ */
+static int
+fs3270_close(struct inode *inode, struct file *filp)
+{
+	struct fs3270 *fp;
+
+	fp = filp->private_data;
+	filp->private_data = NULL;
+	if (fp) {
+		put_pid(fp->fs_pid);
+		fp->fs_pid = NULL;
+		raw3270_reset(&fp->view);
+		raw3270_put_view(&fp->view);
+		raw3270_del_view(&fp->view);
+	}
+	return 0;
+}
+
+static const struct file_operations fs3270_fops = {
+	.owner		 = THIS_MODULE,		/* owner */
+	.read		 = fs3270_read,		/* read */
+	.write		 = fs3270_write,	/* write */
+	.unlocked_ioctl	 = fs3270_ioctl,	/* ioctl */
+	.compat_ioctl	 = fs3270_ioctl,	/* ioctl */
+	.open		 = fs3270_open,		/* open */
+	.release	 = fs3270_close,	/* release */
+	.llseek		= no_llseek,
+};
+
+static void fs3270_create_cb(int minor)
+{
+	__register_chrdev(IBM_FS3270_MAJOR, minor, 1, "tub", &fs3270_fops);
+	device_create(class3270, NULL, MKDEV(IBM_FS3270_MAJOR, minor),
+		      NULL, "3270/tub%d", minor);
+}
+
+static void fs3270_destroy_cb(int minor)
+{
+	device_destroy(class3270, MKDEV(IBM_FS3270_MAJOR, minor));
+	__unregister_chrdev(IBM_FS3270_MAJOR, minor, 1, "tub");
+}
+
+static struct raw3270_notifier fs3270_notifier =
+{
+	.create = fs3270_create_cb,
+	.destroy = fs3270_destroy_cb,
+};
+
+/*
+ * 3270 fullscreen driver initialization.
+ */
+static int __init
+fs3270_init(void)
+{
+	int rc;
+
+	rc = __register_chrdev(IBM_FS3270_MAJOR, 0, 1, "fs3270", &fs3270_fops);
+	if (rc)
+		return rc;
+	device_create(class3270, NULL, MKDEV(IBM_FS3270_MAJOR, 0),
+		      NULL, "3270/tub");
+	raw3270_register_notifier(&fs3270_notifier);
+	return 0;
+}
+
+static void __exit
+fs3270_exit(void)
+{
+	raw3270_unregister_notifier(&fs3270_notifier);
+	device_destroy(class3270, MKDEV(IBM_FS3270_MAJOR, 0));
+	__unregister_chrdev(IBM_FS3270_MAJOR, 0, 1, "fs3270");
+}
+
+MODULE_LICENSE("GPL");
+MODULE_ALIAS_CHARDEV_MAJOR(IBM_FS3270_MAJOR);
+
+module_init(fs3270_init);
+module_exit(fs3270_exit);

drivers/s390/char/hmcdrv_cache.c

@@ -0,0 +1,252 @@
+/*
+ *    SE/HMC Drive (Read) Cache Functions
+ *
+ *    Copyright IBM Corp. 2013
+ *    Author(s): Ralf Hoppe (rhoppe@de.ibm.com)
+ *
+ */
+
+#define KMSG_COMPONENT "hmcdrv"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/jiffies.h>
+
+#include "hmcdrv_ftp.h"
+#include "hmcdrv_cache.h"
+
+#define HMCDRV_CACHE_TIMEOUT		30 /* aging timeout in seconds */
+
+/**
+ * struct hmcdrv_cache_entry - file cache (only used on read/dir)
+ * @id: FTP command ID
+ * @content: kernel-space buffer, 4k aligned
+ * @len: size of @content cache (0 if caching disabled)
+ * @ofs: start of content within file (-1 if no cached content)
+ * @fname: file name
+ * @fsize: file size
+ * @timeout: cache timeout in jiffies
+ *
+ * Notice that the first three members (id, fname, fsize) are cached on all
+ * read/dir requests. But content is cached only under some preconditions.
+ * Uncached content is signalled by a negative value of @ofs.
+ */
+struct hmcdrv_cache_entry {
+	enum hmcdrv_ftp_cmdid id;
+	char fname[HMCDRV_FTP_FIDENT_MAX];
+	size_t fsize;
+	loff_t ofs;
+	unsigned long timeout;
+	void *content;
+	size_t len;
+};
+
+static int hmcdrv_cache_order; /* cache allocated page order */
+
+static struct hmcdrv_cache_entry hmcdrv_cache_file = {
+	.fsize = SIZE_MAX,
+	.ofs = -1,
+	.len = 0,
+	.fname = {'\0'}
+};
+
+/**
+ * hmcdrv_cache_get() - looks for file data/content in read cache
+ * @ftp: pointer to FTP command specification
+ *
+ * Return: number of bytes read from cache or a negative number if nothing
+ * in content cache (for the file/cmd specified in @ftp)
+ */
+static ssize_t hmcdrv_cache_get(const struct hmcdrv_ftp_cmdspec *ftp)
+{
+	loff_t pos; /* position in cache (signed) */
+	ssize_t len;
+
+	if ((ftp->id != hmcdrv_cache_file.id) ||
+	    strcmp(hmcdrv_cache_file.fname, ftp->fname))
+		return -1;
+
+	if (ftp->ofs >= hmcdrv_cache_file.fsize) /* EOF ? */
+		return 0;
+
+	if ((hmcdrv_cache_file.ofs < 0) || /* has content? */
+	    time_after(jiffies, hmcdrv_cache_file.timeout))
+		return -1;
+
+	/* there seems to be cached content - calculate the maximum number
+	 * of bytes that can be returned (regarding file size and offset)
+	 */
+	len = hmcdrv_cache_file.fsize - ftp->ofs;
+
+	if (len > ftp->len)
+		len = ftp->len;
+
+	/* check if the requested chunk falls into our cache (which starts
+	 * at offset 'hmcdrv_cache_file.ofs' in the file of interest)
+	 */
+	pos = ftp->ofs - hmcdrv_cache_file.ofs;
+
+	if ((pos >= 0) &&
+	    ((pos + len) <= hmcdrv_cache_file.len)) {
+
+		memcpy(ftp->buf,
+		       hmcdrv_cache_file.content + pos,
+		       len);
+		pr_debug("using cached content of '%s', returning %zd/%zd bytes\n",
+			 hmcdrv_cache_file.fname, len,
+			 hmcdrv_cache_file.fsize);
+
+		return len;
+	}
+
+	return -1;
+}
+
+/**
+ * hmcdrv_cache_do() - do a HMC drive CD/DVD transfer with cache update
+ * @ftp: pointer to FTP command specification
+ * @func: FTP transfer function to be used
+ *
+ * Return: number of bytes read/written or a (negative) error code
+ */
+static ssize_t hmcdrv_cache_do(const struct hmcdrv_ftp_cmdspec *ftp,
+			       hmcdrv_cache_ftpfunc func)
+{
+	ssize_t len;
+
+	/* only cache content if the read/dir cache really exists
+	 * (hmcdrv_cache_file.len > 0), is large enough to handle the
+	 * request (hmcdrv_cache_file.len >= ftp->len) and there is a need
+	 * to do so (ftp->len > 0)
+	 */
+	if ((ftp->len > 0) && (hmcdrv_cache_file.len >= ftp->len)) {
+
+		/* because the cache is not located at ftp->buf, we have to
+		 * assemble a new HMC drive FTP cmd specification (pointing
+		 * to our cache, and using the increased size)
+		 */
+		struct hmcdrv_ftp_cmdspec cftp = *ftp; /* make a copy */
+		cftp.buf = hmcdrv_cache_file.content;  /* and update */
+		cftp.len = hmcdrv_cache_file.len;      /* buffer data */
+
+		len = func(&cftp, &hmcdrv_cache_file.fsize); /* now do */
+
+		if (len > 0) {
+			pr_debug("caching %zd bytes content for '%s'\n",
+				 len, ftp->fname);
+
+			if (len > ftp->len)
+				len = ftp->len;
+
+			hmcdrv_cache_file.ofs = ftp->ofs;
+			hmcdrv_cache_file.timeout = jiffies +
+				HMCDRV_CACHE_TIMEOUT * HZ;
+			memcpy(ftp->buf, hmcdrv_cache_file.content, len);
+		}
+	} else {
+		len = func(ftp, &hmcdrv_cache_file.fsize);
+		hmcdrv_cache_file.ofs = -1; /* invalidate content */
+	}
+
+	if (len > 0) {
+		/* cache some file info (FTP command, file name and file
+		 * size) unconditionally
+		 */
+		strlcpy(hmcdrv_cache_file.fname, ftp->fname,
+			HMCDRV_FTP_FIDENT_MAX);
+		hmcdrv_cache_file.id = ftp->id;
+		pr_debug("caching cmd %d, file size %zu for '%s'\n",
+			 ftp->id, hmcdrv_cache_file.fsize, ftp->fname);
+	}
+
+	return len;
+}
+
+/**
+ * hmcdrv_cache_cmd() - perform a cached HMC drive CD/DVD transfer
+ * @ftp: pointer to FTP command specification
+ * @func: FTP transfer function to be used
+ *
+ * Attention: Notice that this function is not reentrant - so the caller
+ * must ensure exclusive execution.
+ *
+ * Return: number of bytes read/written or a (negative) error code
+ */
+ssize_t hmcdrv_cache_cmd(const struct hmcdrv_ftp_cmdspec *ftp,
+			 hmcdrv_cache_ftpfunc func)
+{
+	ssize_t len;
+
+	if ((ftp->id == HMCDRV_FTP_DIR) || /* read cache */
+	    (ftp->id == HMCDRV_FTP_NLIST) ||
+	    (ftp->id == HMCDRV_FTP_GET)) {
+
+		len = hmcdrv_cache_get(ftp);
+
+		if (len >= 0) /* got it from cache ? */
+			return len; /* yes */
+
+		len = hmcdrv_cache_do(ftp, func);
+
+		if (len >= 0)
+			return len;
+
+	} else {
+		len = func(ftp, NULL); /* simply do original command */
+	}
+
+	/* invalidate the (read) cache in case there was a write operation
+	 * or an error on read/dir
+	 */
+	hmcdrv_cache_file.id = HMCDRV_FTP_NOOP;
+	hmcdrv_cache_file.fsize = LLONG_MAX;
+	hmcdrv_cache_file.ofs = -1;
+
+	return len;
+}
+
+/**
+ * hmcdrv_cache_startup() - startup of HMC drive cache
+ * @cachesize: cache size
+ *
+ * Return: 0 on success, else a (negative) error code
+ */
+int hmcdrv_cache_startup(size_t cachesize)
+{
+	if (cachesize > 0) { /* perform caching ? */
+		hmcdrv_cache_order = get_order(cachesize);
+		hmcdrv_cache_file.content =
+			(void *) __get_free_pages(GFP_KERNEL | GFP_DMA,
+						  hmcdrv_cache_order);
+
+		if (!hmcdrv_cache_file.content) {
+			pr_err("Allocating the requested cache size of %zu bytes failed\n",
+			       cachesize);
+			return -ENOMEM;
+		}
+
+		pr_debug("content cache enabled, size is %zu bytes\n",
+			 cachesize);
+	}
+
+	hmcdrv_cache_file.len = cachesize;
+	return 0;
+}
+
+/**
+ * hmcdrv_cache_shutdown() - shutdown of HMC drive cache
+ */
+void hmcdrv_cache_shutdown(void)
+{
+	if (hmcdrv_cache_file.content) {
+		free_pages((unsigned long) hmcdrv_cache_file.content,
+			   hmcdrv_cache_order);
+		hmcdrv_cache_file.content = NULL;
+	}
+
+	hmcdrv_cache_file.id = HMCDRV_FTP_NOOP;
+	hmcdrv_cache_file.fsize = LLONG_MAX;
+	hmcdrv_cache_file.ofs = -1;
+	hmcdrv_cache_file.len = 0; /* no cache */
+}

drivers/s390/char/hmcdrv_cache.h

@@ -0,0 +1,24 @@
+/*
+ *    SE/HMC Drive (Read) Cache Functions
+ *
+ *    Copyright IBM Corp. 2013
+ *    Author(s): Ralf Hoppe (rhoppe@de.ibm.com)
+ */
+
+#ifndef __HMCDRV_CACHE_H__
+#define __HMCDRV_CACHE_H__
+
+#include <linux/mmzone.h>
+#include "hmcdrv_ftp.h"
+
+#define HMCDRV_CACHE_SIZE_DFLT	(MAX_ORDER_NR_PAGES * PAGE_SIZE / 2UL)
+
+typedef ssize_t (*hmcdrv_cache_ftpfunc)(const struct hmcdrv_ftp_cmdspec *ftp,
+					size_t *fsize);
+
+ssize_t hmcdrv_cache_cmd(const struct hmcdrv_ftp_cmdspec *ftp,
+			 hmcdrv_cache_ftpfunc func);
+int hmcdrv_cache_startup(size_t cachesize);
+void hmcdrv_cache_shutdown(void);
+
+#endif	 /* __HMCDRV_CACHE_H__ */

drivers/s390/char/hmcdrv_dev.c

@@ -0,0 +1,370 @@
+/*
+ *    HMC Drive CD/DVD Device
+ *
+ *    Copyright IBM Corp. 2013
+ *    Author(s): Ralf Hoppe (rhoppe@de.ibm.com)
+ *
+ *    This file provides a Linux "misc" character device for access to an
+ *    assigned HMC drive CD/DVD-ROM. It works as follows: First create the
+ *    device by calling hmcdrv_dev_init(). After open() a lseek(fd, 0,
+ *    SEEK_END) indicates that a new FTP command follows (not needed on the
+ *    first command after open). Then write() the FTP command ASCII string
+ *    to it, e.g. "dir /" or "nls <directory>" or "get <filename>". At the
+ *    end read() the response.
+ */
+
+#define KMSG_COMPONENT "hmcdrv"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/fs.h>
+#include <linux/cdev.h>
+#include <linux/miscdevice.h>
+#include <linux/device.h>
+#include <linux/capability.h>
+#include <linux/delay.h>
+#include <linux/uaccess.h>
+
+#include "hmcdrv_dev.h"
+#include "hmcdrv_ftp.h"
+
+/* If the following macro is defined, then the HMC device creates it's own
+ * separated device class (and dynamically assigns a major number). If not
+ * defined then the HMC device is assigned to the "misc" class devices.
+ *
+#define HMCDRV_DEV_CLASS "hmcftp"
+ */
+
+#define HMCDRV_DEV_NAME  "hmcdrv"
+#define HMCDRV_DEV_BUSY_DELAY	 500 /* delay between -EBUSY trials in ms */
+#define HMCDRV_DEV_BUSY_RETRIES  3   /* number of retries on -EBUSY */
+
+struct hmcdrv_dev_node {
+
+#ifdef HMCDRV_DEV_CLASS
+	struct cdev dev; /* character device structure */
+	umode_t mode;	 /* mode of device node (unused, zero) */
+#else
+	struct miscdevice dev; /* "misc" device structure */
+#endif
+
+};
+
+static int hmcdrv_dev_open(struct inode *inode, struct file *fp);
+static int hmcdrv_dev_release(struct inode *inode, struct file *fp);
+static loff_t hmcdrv_dev_seek(struct file *fp, loff_t pos, int whence);
+static ssize_t hmcdrv_dev_read(struct file *fp, char __user *ubuf,
+			       size_t len, loff_t *pos);
+static ssize_t hmcdrv_dev_write(struct file *fp, const char __user *ubuf,
+				size_t len, loff_t *pos);
+static ssize_t hmcdrv_dev_transfer(char __kernel *cmd, loff_t offset,
+				   char __user *buf, size_t len);
+
+/*
+ * device operations
+ */
+static const struct file_operations hmcdrv_dev_fops = {
+	.open = hmcdrv_dev_open,
+	.llseek = hmcdrv_dev_seek,
+	.release = hmcdrv_dev_release,
+	.read = hmcdrv_dev_read,
+	.write = hmcdrv_dev_write,
+};
+
+static struct hmcdrv_dev_node hmcdrv_dev; /* HMC device struct (static) */
+
+#ifdef HMCDRV_DEV_CLASS
+
+static struct class *hmcdrv_dev_class; /* device class pointer */
+static dev_t hmcdrv_dev_no; /* device number (major/minor) */
+
+/**
+ * hmcdrv_dev_name() - provides a naming hint for a device node in /dev
+ * @dev: device for which the naming/mode hint is
+ * @mode: file mode for device node created in /dev
+ *
+ * See: devtmpfs.c, function devtmpfs_create_node()
+ *
+ * Return: recommended device file name in /dev
+ */
+static char *hmcdrv_dev_name(struct device *dev, umode_t *mode)
+{
+	char *nodename = NULL;
+	const char *devname = dev_name(dev); /* kernel device name */
+
+	if (devname)
+		nodename = kasprintf(GFP_KERNEL, "%s", devname);
+
+	/* on device destroy (rmmod) the mode pointer may be NULL
+	 */
+	if (mode)
+		*mode = hmcdrv_dev.mode;
+
+	return nodename;
+}
+
+#endif	/* HMCDRV_DEV_CLASS */
+
+/*
+ * open()
+ */
+static int hmcdrv_dev_open(struct inode *inode, struct file *fp)
+{
+	int rc;
+
+	/* check for non-blocking access, which is really unsupported
+	 */
+	if (fp->f_flags & O_NONBLOCK)
+		return -EINVAL;
+
+	/* Because it makes no sense to open this device read-only (then a
+	 * FTP command cannot be emitted), we respond with an error.
+	 */
+	if ((fp->f_flags & O_ACCMODE) == O_RDONLY)
+		return -EINVAL;
+
+	/* prevent unloading this module as long as anyone holds the
+	 * device file open - so increment the reference count here
+	 */
+	if (!try_module_get(THIS_MODULE))
+		return -ENODEV;
+
+	fp->private_data = NULL; /* no command yet */
+	rc = hmcdrv_ftp_startup();
+	if (rc)
+		module_put(THIS_MODULE);
+
+	pr_debug("open file '/dev/%s' with return code %d\n",
+		 fp->f_dentry->d_name.name, rc);
+	return rc;
+}
+
+/*
+ * release()
+ */
+static int hmcdrv_dev_release(struct inode *inode, struct file *fp)
+{
+	pr_debug("closing file '/dev/%s'\n", fp->f_dentry->d_name.name);
+	kfree(fp->private_data);
+	fp->private_data = NULL;
+	hmcdrv_ftp_shutdown();
+	module_put(THIS_MODULE);
+	return 0;
+}
+
+/*
+ * lseek()
+ */
+static loff_t hmcdrv_dev_seek(struct file *fp, loff_t pos, int whence)
+{
+	switch (whence) {
+	case SEEK_CUR: /* relative to current file position */
+		pos += fp->f_pos; /* new position stored in 'pos' */
+		break;
+
+	case SEEK_SET: /* absolute (relative to beginning of file) */
+		break; /* SEEK_SET */
+
+		/* We use SEEK_END as a special indicator for a SEEK_SET
+		 * (set absolute position), combined with a FTP command
+		 * clear.
+		 */
+	case SEEK_END:
+		if (fp->private_data) {
+			kfree(fp->private_data);
+			fp->private_data = NULL;
+		}
+
+		break; /* SEEK_END */
+
+	default: /* SEEK_DATA, SEEK_HOLE: unsupported */
+		return -EINVAL;
+	}
+
+	if (pos < 0)
+		return -EINVAL;
+
+	if (fp->f_pos != pos)
+		++fp->f_version;
+
+	fp->f_pos = pos;
+	return pos;
+}
+
+/*
+ * transfer (helper function)
+ */
+static ssize_t hmcdrv_dev_transfer(char __kernel *cmd, loff_t offset,
+				   char __user *buf, size_t len)
+{
+	ssize_t retlen;
+	unsigned trials = HMCDRV_DEV_BUSY_RETRIES;
+
+	do {
+		retlen = hmcdrv_ftp_cmd(cmd, offset, buf, len);
+
+		if (retlen != -EBUSY)
+			break;
+
+		msleep(HMCDRV_DEV_BUSY_DELAY);
+
+	} while (--trials > 0);
+
+	return retlen;
+}
+
+/*
+ * read()
+ */
+static ssize_t hmcdrv_dev_read(struct file *fp, char __user *ubuf,
+			       size_t len, loff_t *pos)
+{
+	ssize_t retlen;
+
+	if (((fp->f_flags & O_ACCMODE) == O_WRONLY) ||
+	    (fp->private_data == NULL)) { /* no FTP cmd defined ? */
+		return -EBADF;
+	}
+
+	retlen = hmcdrv_dev_transfer((char *) fp->private_data,
+				     *pos, ubuf, len);
+
+	pr_debug("read from file '/dev/%s' at %lld returns %zd/%zu\n",
+		 fp->f_dentry->d_name.name, (long long) *pos, retlen, len);
+
+	if (retlen > 0)
+		*pos += retlen;
+
+	return retlen;
+}
+
+/*
+ * write()
+ */
+static ssize_t hmcdrv_dev_write(struct file *fp, const char __user *ubuf,
+				size_t len, loff_t *pos)
+{
+	ssize_t retlen;
+
+	pr_debug("writing file '/dev/%s' at pos. %lld with length %zd\n",
+		 fp->f_dentry->d_name.name, (long long) *pos, len);
+
+	if (!fp->private_data) { /* first expect a cmd write */
+		fp->private_data = kmalloc(len + 1, GFP_KERNEL);
+
+		if (!fp->private_data)
+			return -ENOMEM;
+
+		if (!copy_from_user(fp->private_data, ubuf, len)) {
+			((char *)fp->private_data)[len] = '\0';
+			return len;
+		}
+
+		kfree(fp->private_data);
+		fp->private_data = NULL;
+		return -EFAULT;
+	}
+
+	retlen = hmcdrv_dev_transfer((char *) fp->private_data,
+				     *pos, (char __user *) ubuf, len);
+	if (retlen > 0)
+		*pos += retlen;
+
+	pr_debug("write to file '/dev/%s' returned %zd\n",
+		 fp->f_dentry->d_name.name, retlen);
+
+	return retlen;
+}
+
+/**
+ * hmcdrv_dev_init() - creates a HMC drive CD/DVD device
+ *
+ * This function creates a HMC drive CD/DVD kernel device and an associated
+ * device under /dev, using a dynamically allocated major number.
+ *
+ * Return: 0 on success, else an error code.
+ */
+int hmcdrv_dev_init(void)
+{
+	int rc;
+
+#ifdef HMCDRV_DEV_CLASS
+	struct device *dev;
+
+	rc = alloc_chrdev_region(&hmcdrv_dev_no, 0, 1, HMCDRV_DEV_NAME);
+
+	if (rc)
+		goto out_err;
+
+	cdev_init(&hmcdrv_dev.dev, &hmcdrv_dev_fops);
+	hmcdrv_dev.dev.owner = THIS_MODULE;
+	rc = cdev_add(&hmcdrv_dev.dev, hmcdrv_dev_no, 1);
+
+	if (rc)
+		goto out_unreg;
+
+	/* At this point the character device exists in the kernel (see
+	 * /proc/devices), but not under /dev nor /sys/devices/virtual. So
+	 * we have to create an associated class (see /sys/class).
+	 */
+	hmcdrv_dev_class = class_create(THIS_MODULE, HMCDRV_DEV_CLASS);
+
+	if (IS_ERR(hmcdrv_dev_class)) {
+		rc = PTR_ERR(hmcdrv_dev_class);
+		goto out_devdel;
+	}
+
+	/* Finally a device node in /dev has to be established (as 'mkdev'
+	 * does from the command line). Notice that assignment of a device
+	 * node name/mode function is optional (only for mode != 0600).
+	 */
+	hmcdrv_dev.mode = 0; /* "unset" */
+	hmcdrv_dev_class->devnode = hmcdrv_dev_name;
+
+	dev = device_create(hmcdrv_dev_class, NULL, hmcdrv_dev_no, NULL,
+			    "%s", HMCDRV_DEV_NAME);
+	if (!IS_ERR(dev))
+		return 0;
+
+	rc = PTR_ERR(dev);
+	class_destroy(hmcdrv_dev_class);
+	hmcdrv_dev_class = NULL;
+
+out_devdel:
+	cdev_del(&hmcdrv_dev.dev);
+
+out_unreg:
+	unregister_chrdev_region(hmcdrv_dev_no, 1);
+
+out_err:
+
+#else  /* !HMCDRV_DEV_CLASS */
+	hmcdrv_dev.dev.minor = MISC_DYNAMIC_MINOR;
+	hmcdrv_dev.dev.name = HMCDRV_DEV_NAME;
+	hmcdrv_dev.dev.fops = &hmcdrv_dev_fops;
+	hmcdrv_dev.dev.mode = 0; /* finally produces 0600 */
+	rc = misc_register(&hmcdrv_dev.dev);
+#endif	/* HMCDRV_DEV_CLASS */
+
+	return rc;
+}
+
+/**
+ * hmcdrv_dev_exit() - destroys a HMC drive CD/DVD device
+ */
+void hmcdrv_dev_exit(void)
+{
+#ifdef HMCDRV_DEV_CLASS
+	if (!IS_ERR_OR_NULL(hmcdrv_dev_class)) {
+		device_destroy(hmcdrv_dev_class, hmcdrv_dev_no);
+		class_destroy(hmcdrv_dev_class);
+	}
+
+	cdev_del(&hmcdrv_dev.dev);
+	unregister_chrdev_region(hmcdrv_dev_no, 1);
+#else  /* !HMCDRV_DEV_CLASS */
+	misc_deregister(&hmcdrv_dev.dev);
+#endif	/* HMCDRV_DEV_CLASS */
+}

drivers/s390/char/hmcdrv_dev.h

@@ -0,0 +1,14 @@
+/*
+ *    SE/HMC Drive FTP Device
+ *
+ *    Copyright IBM Corp. 2013
+ *    Author(s): Ralf Hoppe (rhoppe@de.ibm.com)
+ */
+
+#ifndef __HMCDRV_DEV_H__
+#define __HMCDRV_DEV_H__
+
+int hmcdrv_dev_init(void);
+void hmcdrv_dev_exit(void);
+
+#endif	 /* __HMCDRV_DEV_H__ */

drivers/s390/char/hmcdrv_ftp.c

@@ -0,0 +1,343 @@
+/*
+ *    HMC Drive FTP Services
+ *
+ *    Copyright IBM Corp. 2013
+ *    Author(s): Ralf Hoppe (rhoppe@de.ibm.com)
+ */
+
+#define KMSG_COMPONENT "hmcdrv"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/uaccess.h>
+#include <linux/export.h>
+
+#include <linux/ctype.h>
+#include <linux/crc16.h>
+
+#include "hmcdrv_ftp.h"
+#include "hmcdrv_cache.h"
+#include "sclp_ftp.h"
+#include "diag_ftp.h"
+
+/**
+ * struct hmcdrv_ftp_ops - HMC drive FTP operations
+ * @startup: startup function
+ * @shutdown: shutdown function
+ * @cmd: FTP transfer function
+ */
+struct hmcdrv_ftp_ops {
+	int (*startup)(void);
+	void (*shutdown)(void);
+	ssize_t (*transfer)(const struct hmcdrv_ftp_cmdspec *ftp,
+			    size_t *fsize);
+};
+
+static enum hmcdrv_ftp_cmdid hmcdrv_ftp_cmd_getid(const char *cmd, int len);
+static int hmcdrv_ftp_parse(char *cmd, struct hmcdrv_ftp_cmdspec *ftp);
+
+static struct hmcdrv_ftp_ops *hmcdrv_ftp_funcs; /* current operations */
+static DEFINE_MUTEX(hmcdrv_ftp_mutex); /* mutex for hmcdrv_ftp_funcs */
+static unsigned hmcdrv_ftp_refcnt; /* start/shutdown reference counter */
+
+/**
+ * hmcdrv_ftp_cmd_getid() - determine FTP command ID from a command string
+ * @cmd: FTP command string (NOT zero-terminated)
+ * @len: length of FTP command string in @cmd
+ */
+static enum hmcdrv_ftp_cmdid hmcdrv_ftp_cmd_getid(const char *cmd, int len)
+{
+	/* HMC FTP command descriptor */
+	struct hmcdrv_ftp_cmd_desc {
+		const char *str;	   /* command string */
+		enum hmcdrv_ftp_cmdid cmd; /* associated command as enum */
+	};
+
+	/* Description of all HMC drive FTP commands
+	 *
+	 * Notes:
+	 * 1. Array size should be a prime number.
+	 * 2. Do not change the order of commands in table (because the
+	 *    index is determined by CRC % ARRAY_SIZE).
+	 * 3. Original command 'nlist' was renamed, else the CRC would
+	 *    collide with 'append' (see point 2).
+	 */
+	static const struct hmcdrv_ftp_cmd_desc ftpcmds[7] = {
+
+		{.str = "get", /* [0] get (CRC = 0x68eb) */
+		 .cmd = HMCDRV_FTP_GET},
+		{.str = "dir", /* [1] dir (CRC = 0x6a9e) */
+		 .cmd = HMCDRV_FTP_DIR},
+		{.str = "delete", /* [2] delete (CRC = 0x53ae) */
+		 .cmd = HMCDRV_FTP_DELETE},
+		{.str = "nls", /* [3] nls (CRC = 0xf87c) */
+		 .cmd = HMCDRV_FTP_NLIST},
+		{.str = "put", /* [4] put (CRC = 0xac56) */
+		 .cmd = HMCDRV_FTP_PUT},
+		{.str = "append", /* [5] append (CRC = 0xf56e) */
+		 .cmd = HMCDRV_FTP_APPEND},
+		{.str = NULL} /* [6] unused */
+	};
+
+	const struct hmcdrv_ftp_cmd_desc *pdesc;
+
+	u16 crc = 0xffffU;
+
+	if (len == 0)
+		return HMCDRV_FTP_NOOP; /* error indiactor */
+
+	crc = crc16(crc, cmd, len);
+	pdesc = ftpcmds + (crc % ARRAY_SIZE(ftpcmds));
+	pr_debug("FTP command '%s' has CRC 0x%04x, at table pos. %lu\n",
+		 cmd, crc, (crc % ARRAY_SIZE(ftpcmds)));
+
+	if (!pdesc->str || strncmp(pdesc->str, cmd, len))
+		return HMCDRV_FTP_NOOP;
+
+	pr_debug("FTP command '%s' found, with ID %d\n",
+		 pdesc->str, pdesc->cmd);
+
+	return pdesc->cmd;
+}
+
+/**
+ * hmcdrv_ftp_parse() - HMC drive FTP command parser
+ * @cmd: FTP command string "<cmd> <filename>"
+ * @ftp: Pointer to FTP command specification buffer (output)
+ *
+ * Return: 0 on success, else a (negative) error code
+ */
+static int hmcdrv_ftp_parse(char *cmd, struct hmcdrv_ftp_cmdspec *ftp)
+{
+	char *start;
+	int argc = 0;
+
+	ftp->id = HMCDRV_FTP_NOOP;
+	ftp->fname = NULL;
+
+	while (*cmd != '\0') {
+
+		while (isspace(*cmd))
+			++cmd;
+
+		if (*cmd == '\0')
+			break;
+
+		start = cmd;
+
+		switch (argc) {
+		case 0: /* 1st argument (FTP command) */
+			while ((*cmd != '\0') && !isspace(*cmd))
+				++cmd;
+			ftp->id = hmcdrv_ftp_cmd_getid(start, cmd - start);
+			break;
+		case 1: /* 2nd / last argument (rest of line) */
+			while ((*cmd != '\0') && !iscntrl(*cmd))
+				++cmd;
+			ftp->fname = start;
+			/* fall through */
+		default:
+			*cmd = '\0';
+			break;
+		} /* switch */
+
+		++argc;
+	} /* while */
+
+	if (!ftp->fname || (ftp->id == HMCDRV_FTP_NOOP))
+		return -EINVAL;
+
+	return 0;
+}
+
+/**
+ * hmcdrv_ftp_do() - perform a HMC drive FTP, with data from kernel-space
+ * @ftp: pointer to FTP command specification
+ *
+ * Return: number of bytes read/written or a negative error code
+ */
+ssize_t hmcdrv_ftp_do(const struct hmcdrv_ftp_cmdspec *ftp)
+{
+	ssize_t len;
+
+	mutex_lock(&hmcdrv_ftp_mutex);
+
+	if (hmcdrv_ftp_funcs && hmcdrv_ftp_refcnt) {
+		pr_debug("starting transfer, cmd %d for '%s' at %lld with %zd bytes\n",
+			 ftp->id, ftp->fname, (long long) ftp->ofs, ftp->len);
+		len = hmcdrv_cache_cmd(ftp, hmcdrv_ftp_funcs->transfer);
+	} else {
+		len = -ENXIO;
+	}
+
+	mutex_unlock(&hmcdrv_ftp_mutex);
+	return len;
+}
+EXPORT_SYMBOL(hmcdrv_ftp_do);
+
+/**
+ * hmcdrv_ftp_probe() - probe for the HMC drive FTP service
+ *
+ * Return: 0 if service is available, else an (negative) error code
+ */
+int hmcdrv_ftp_probe(void)
+{
+	int rc;
+
+	struct hmcdrv_ftp_cmdspec ftp = {
+		.id = HMCDRV_FTP_NOOP,
+		.ofs = 0,
+		.fname = "",
+		.len = PAGE_SIZE
+	};
+
+	ftp.buf = (void *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+
+	if (!ftp.buf)
+		return -ENOMEM;
+
+	rc = hmcdrv_ftp_startup();
+
+	if (rc)
+		return rc;
+
+	rc = hmcdrv_ftp_do(&ftp);
+	free_page((unsigned long) ftp.buf);
+	hmcdrv_ftp_shutdown();
+
+	switch (rc) {
+	case -ENOENT: /* no such file/media or currently busy, */
+	case -EBUSY:  /* but service seems to be available */
+		rc = 0;
+		break;
+	default: /* leave 'rc' as it is for [0, -EPERM, -E...] */
+		if (rc > 0)
+			rc = 0; /* clear length (success) */
+		break;
+	} /* switch */
+
+	return rc;
+}
+EXPORT_SYMBOL(hmcdrv_ftp_probe);
+
+/**
+ * hmcdrv_ftp_cmd() - Perform a HMC drive FTP, with data from user-space
+ *
+ * @cmd: FTP command string "<cmd> <filename>"
+ * @offset: file position to read/write
+ * @buf: user-space buffer for read/written directory/file
+ * @len: size of @buf (read/dir) or number of bytes to write
+ *
+ * This function must not be called before hmcdrv_ftp_startup() was called.
+ *
+ * Return: number of bytes read/written or a negative error code
+ */
+ssize_t hmcdrv_ftp_cmd(char __kernel *cmd, loff_t offset,
+		       char __user *buf, size_t len)
+{
+	int order;
+
+	struct hmcdrv_ftp_cmdspec ftp = {.len = len, .ofs = offset};
+	ssize_t retlen = hmcdrv_ftp_parse(cmd, &ftp);
+
+	if (retlen)
+		return retlen;
+
+	order = get_order(ftp.len);
+	ftp.buf = (void *) __get_free_pages(GFP_KERNEL | GFP_DMA, order);
+
+	if (!ftp.buf)
+		return -ENOMEM;
+
+	switch (ftp.id) {
+	case HMCDRV_FTP_DIR:
+	case HMCDRV_FTP_NLIST:
+	case HMCDRV_FTP_GET:
+		retlen = hmcdrv_ftp_do(&ftp);
+
+		if ((retlen >= 0) &&
+		    copy_to_user(buf, ftp.buf, retlen))
+			retlen = -EFAULT;
+		break;
+
+	case HMCDRV_FTP_PUT:
+	case HMCDRV_FTP_APPEND:
+		if (!copy_from_user(ftp.buf, buf, ftp.len))
+			retlen = hmcdrv_ftp_do(&ftp);
+		else
+			retlen = -EFAULT;
+		break;
+
+	case HMCDRV_FTP_DELETE:
+		retlen = hmcdrv_ftp_do(&ftp);
+		break;
+
+	default:
+		retlen = -EOPNOTSUPP;
+		break;
+	}
+
+	free_pages((unsigned long) ftp.buf, order);
+	return retlen;
+}
+
+/**
+ * hmcdrv_ftp_startup() - startup of HMC drive FTP functionality for a
+ * dedicated (owner) instance
+ *
+ * Return: 0 on success, else an (negative) error code
+ */
+int hmcdrv_ftp_startup(void)
+{
+	static struct hmcdrv_ftp_ops hmcdrv_ftp_zvm = {
+		.startup = diag_ftp_startup,
+		.shutdown = diag_ftp_shutdown,
+		.transfer = diag_ftp_cmd
+	};
+
+	static struct hmcdrv_ftp_ops hmcdrv_ftp_lpar = {
+		.startup = sclp_ftp_startup,
+		.shutdown = sclp_ftp_shutdown,
+		.transfer = sclp_ftp_cmd
+	};
+
+	int rc = 0;
+
+	mutex_lock(&hmcdrv_ftp_mutex); /* block transfers while start-up */
+
+	if (hmcdrv_ftp_refcnt == 0) {
+		if (MACHINE_IS_VM)
+			hmcdrv_ftp_funcs = &hmcdrv_ftp_zvm;
+		else if (MACHINE_IS_LPAR || MACHINE_IS_KVM)
+			hmcdrv_ftp_funcs = &hmcdrv_ftp_lpar;
+		else
+			rc = -EOPNOTSUPP;
+
+		if (hmcdrv_ftp_funcs)
+			rc = hmcdrv_ftp_funcs->startup();
+	}
+
+	if (!rc)
+		++hmcdrv_ftp_refcnt;
+
+	mutex_unlock(&hmcdrv_ftp_mutex);
+	return rc;
+}
+EXPORT_SYMBOL(hmcdrv_ftp_startup);
+
+/**
+ * hmcdrv_ftp_shutdown() - shutdown of HMC drive FTP functionality for a
+ * dedicated (owner) instance
+ */
+void hmcdrv_ftp_shutdown(void)
+{
+	mutex_lock(&hmcdrv_ftp_mutex);
+	--hmcdrv_ftp_refcnt;
+
+	if ((hmcdrv_ftp_refcnt == 0) && hmcdrv_ftp_funcs)
+		hmcdrv_ftp_funcs->shutdown();
+
+	mutex_unlock(&hmcdrv_ftp_mutex);
+}
+EXPORT_SYMBOL(hmcdrv_ftp_shutdown);

drivers/s390/char/hmcdrv_ftp.h

@@ -0,0 +1,63 @@
+/*
+ *    SE/HMC Drive FTP Services
+ *
+ *    Copyright IBM Corp. 2013
+ *    Author(s): Ralf Hoppe (rhoppe@de.ibm.com)
+ */
+
+#ifndef __HMCDRV_FTP_H__
+#define __HMCDRV_FTP_H__
+
+#include <linux/types.h> /* size_t, loff_t */
+
+/*
+ * HMC drive FTP Service max. length of path (w/ EOS)
+ */
+#define HMCDRV_FTP_FIDENT_MAX 192
+
+/**
+ * enum hmcdrv_ftp_cmdid - HMC drive FTP commands
+ * @HMCDRV_FTP_NOOP: do nothing (only for probing)
+ * @HMCDRV_FTP_GET: read a file
+ * @HMCDRV_FTP_PUT: (over-) write a file
+ * @HMCDRV_FTP_APPEND: append to a file
+ * @HMCDRV_FTP_DIR: list directory long (ls -l)
+ * @HMCDRV_FTP_NLIST: list files, no directories (name list)
+ * @HMCDRV_FTP_DELETE: delete a file
+ * @HMCDRV_FTP_CANCEL: cancel operation (SCLP/LPAR only)
+ */
+enum hmcdrv_ftp_cmdid {
+	HMCDRV_FTP_NOOP = 0,
+	HMCDRV_FTP_GET = 1,
+	HMCDRV_FTP_PUT = 2,
+	HMCDRV_FTP_APPEND = 3,
+	HMCDRV_FTP_DIR = 4,
+	HMCDRV_FTP_NLIST = 5,
+	HMCDRV_FTP_DELETE = 6,
+	HMCDRV_FTP_CANCEL = 7
+};
+
+/**
+ * struct hmcdrv_ftp_cmdspec - FTP command specification
+ * @id: FTP command ID
+ * @ofs: offset in file
+ * @fname: filename (ASCII), null-terminated
+ * @buf: kernel-space transfer data buffer, 4k aligned
+ * @len: (max) number of bytes to transfer from/to @buf
+ */
+struct hmcdrv_ftp_cmdspec {
+	enum hmcdrv_ftp_cmdid id;
+	loff_t ofs;
+	const char *fname;
+	void __kernel *buf;
+	size_t len;
+};
+
+int hmcdrv_ftp_startup(void);
+void hmcdrv_ftp_shutdown(void);
+int hmcdrv_ftp_probe(void);
+ssize_t hmcdrv_ftp_do(const struct hmcdrv_ftp_cmdspec *ftp);
+ssize_t hmcdrv_ftp_cmd(char __kernel *cmd, loff_t offset,
+		       char __user *buf, size_t len);
+
+#endif	 /* __HMCDRV_FTP_H__ */

drivers/s390/char/hmcdrv_mod.c

@@ -0,0 +1,64 @@
+/*
+ *    HMC Drive DVD Module
+ *
+ *    Copyright IBM Corp. 2013
+ *    Author(s): Ralf Hoppe (rhoppe@de.ibm.com)
+ */
+
+#define KMSG_COMPONENT "hmcdrv"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/version.h>
+#include <linux/stat.h>
+
+#include "hmcdrv_ftp.h"
+#include "hmcdrv_dev.h"
+#include "hmcdrv_cache.h"
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Copyright 2013 IBM Corporation");
+MODULE_DESCRIPTION("HMC drive DVD access");
+
+/*
+ * module parameter 'cachesize'
+ */
+static size_t hmcdrv_mod_cachesize = HMCDRV_CACHE_SIZE_DFLT;
+module_param_named(cachesize, hmcdrv_mod_cachesize, ulong, S_IRUGO);
+
+/**
+ * hmcdrv_mod_init() - module init function
+ */
+static int __init hmcdrv_mod_init(void)
+{
+	int rc = hmcdrv_ftp_probe(); /* perform w/o cache */
+
+	if (rc)
+		return rc;
+
+	rc = hmcdrv_cache_startup(hmcdrv_mod_cachesize);
+
+	if (rc)
+		return rc;
+
+	rc = hmcdrv_dev_init();
+
+	if (rc)
+		hmcdrv_cache_shutdown();
+
+	return rc;
+}
+
+/**
+ * hmcdrv_mod_exit() - module exit function
+ */
+static void __exit hmcdrv_mod_exit(void)
+{
+	hmcdrv_dev_exit();
+	hmcdrv_cache_shutdown();
+}
+
+module_init(hmcdrv_mod_init);
+module_exit(hmcdrv_mod_exit);

drivers/s390/char/keyboard.c

@@ -0,0 +1,559 @@
+/*
+ *    ebcdic keycode functions for s390 console drivers
+ *
+ *  S390 version
+ *    Copyright IBM Corp. 2003
+ *    Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com),
+ */
+
+#include <linux/module.h>
+#include <linux/sched.h>
+#include <linux/slab.h>
+#include <linux/sysrq.h>
+
+#include <linux/consolemap.h>
+#include <linux/kbd_kern.h>
+#include <linux/kbd_diacr.h>
+#include <asm/uaccess.h>
+
+#include "keyboard.h"
+
+/*
+ * Handler Tables.
+ */
+#define K_HANDLERS\
+	k_self,		k_fn,		k_spec,		k_ignore,\
+	k_dead,		k_ignore,	k_ignore,	k_ignore,\
+	k_ignore,	k_ignore,	k_ignore,	k_ignore,\
+	k_ignore,	k_ignore,	k_ignore,	k_ignore
+
+typedef void (k_handler_fn)(struct kbd_data *, unsigned char);
+static k_handler_fn K_HANDLERS;
+static k_handler_fn *k_handler[16] = { K_HANDLERS };
+
+/* maximum values each key_handler can handle */
+static const int kbd_max_vals[] = {
+	255, ARRAY_SIZE(func_table) - 1, NR_FN_HANDLER - 1, 0,
+	NR_DEAD - 1, 0, 0, 0, 0, 0, 0, 0, 0, 0
+};
+static const int KBD_NR_TYPES = ARRAY_SIZE(kbd_max_vals);
+
+static unsigned char ret_diacr[NR_DEAD] = {
+	'`', '\'', '^', '~', '"', ','
+};
+
+/*
+ * Alloc/free of kbd_data structures.
+ */
+struct kbd_data *
+kbd_alloc(void) {
+	struct kbd_data *kbd;
+	int i;
+
+	kbd = kzalloc(sizeof(struct kbd_data), GFP_KERNEL);
+	if (!kbd)
+		goto out;
+	kbd->key_maps = kzalloc(sizeof(key_maps), GFP_KERNEL);
+	if (!kbd->key_maps)
+		goto out_kbd;
+	for (i = 0; i < ARRAY_SIZE(key_maps); i++) {
+		if (key_maps[i]) {
+			kbd->key_maps[i] = kmemdup(key_maps[i],
+						   sizeof(u_short) * NR_KEYS,
+						   GFP_KERNEL);
+			if (!kbd->key_maps[i])
+				goto out_maps;
+		}
+	}
+	kbd->func_table = kzalloc(sizeof(func_table), GFP_KERNEL);
+	if (!kbd->func_table)
+		goto out_maps;
+	for (i = 0; i < ARRAY_SIZE(func_table); i++) {
+		if (func_table[i]) {
+			kbd->func_table[i] = kstrdup(func_table[i],
+						     GFP_KERNEL);
+			if (!kbd->func_table[i])
+				goto out_func;
+		}
+	}
+	kbd->fn_handler =
+		kzalloc(sizeof(fn_handler_fn *) * NR_FN_HANDLER, GFP_KERNEL);
+	if (!kbd->fn_handler)
+		goto out_func;
+	kbd->accent_table = kmemdup(accent_table,
+				    sizeof(struct kbdiacruc) * MAX_DIACR,
+				    GFP_KERNEL);
+	if (!kbd->accent_table)
+		goto out_fn_handler;
+	kbd->accent_table_size = accent_table_size;
+	return kbd;
+
+out_fn_handler:
+	kfree(kbd->fn_handler);
+out_func:
+	for (i = 0; i < ARRAY_SIZE(func_table); i++)
+		kfree(kbd->func_table[i]);
+	kfree(kbd->func_table);
+out_maps:
+	for (i = 0; i < ARRAY_SIZE(key_maps); i++)
+		kfree(kbd->key_maps[i]);
+	kfree(kbd->key_maps);
+out_kbd:
+	kfree(kbd);
+out:
+	return NULL;
+}
+
+void
+kbd_free(struct kbd_data *kbd)
+{
+	int i;
+
+	kfree(kbd->accent_table);
+	kfree(kbd->fn_handler);
+	for (i = 0; i < ARRAY_SIZE(func_table); i++)
+		kfree(kbd->func_table[i]);
+	kfree(kbd->func_table);
+	for (i = 0; i < ARRAY_SIZE(key_maps); i++)
+		kfree(kbd->key_maps[i]);
+	kfree(kbd->key_maps);
+	kfree(kbd);
+}
+
+/*
+ * Generate ascii -> ebcdic translation table from kbd_data.
+ */
+void
+kbd_ascebc(struct kbd_data *kbd, unsigned char *ascebc)
+{
+	unsigned short *keymap, keysym;
+	int i, j, k;
+
+	memset(ascebc, 0x40, 256);
+	for (i = 0; i < ARRAY_SIZE(key_maps); i++) {
+		keymap = kbd->key_maps[i];
+		if (!keymap)
+			continue;
+		for (j = 0; j < NR_KEYS; j++) {
+			k = ((i & 1) << 7) + j;
+			keysym = keymap[j];
+			if (KTYP(keysym) == (KT_LATIN | 0xf0) ||
+			    KTYP(keysym) == (KT_LETTER | 0xf0))
+				ascebc[KVAL(keysym)] = k;
+			else if (KTYP(keysym) == (KT_DEAD | 0xf0))
+				ascebc[ret_diacr[KVAL(keysym)]] = k;
+		}
+	}
+}
+
+#if 0
+/*
+ * Generate ebcdic -> ascii translation table from kbd_data.
+ */
+void
+kbd_ebcasc(struct kbd_data *kbd, unsigned char *ebcasc)
+{
+	unsigned short *keymap, keysym;
+	int i, j, k;
+
+	memset(ebcasc, ' ', 256);
+	for (i = 0; i < ARRAY_SIZE(key_maps); i++) {
+		keymap = kbd->key_maps[i];
+		if (!keymap)
+			continue;
+		for (j = 0; j < NR_KEYS; j++) {
+			keysym = keymap[j];
+			k = ((i & 1) << 7) + j;
+			if (KTYP(keysym) == (KT_LATIN | 0xf0) ||
+			    KTYP(keysym) == (KT_LETTER | 0xf0))
+				ebcasc[k] = KVAL(keysym);
+			else if (KTYP(keysym) == (KT_DEAD | 0xf0))
+				ebcasc[k] = ret_diacr[KVAL(keysym)];
+		}
+	}
+}
+#endif
+
+/*
+ * We have a combining character DIACR here, followed by the character CH.
+ * If the combination occurs in the table, return the corresponding value.
+ * Otherwise, if CH is a space or equals DIACR, return DIACR.
+ * Otherwise, conclude that DIACR was not combining after all,
+ * queue it and return CH.
+ */
+static unsigned int
+handle_diacr(struct kbd_data *kbd, unsigned int ch)
+{
+	int i, d;
+
+	d = kbd->diacr;
+	kbd->diacr = 0;
+
+	for (i = 0; i < kbd->accent_table_size; i++) {
+		if (kbd->accent_table[i].diacr == d &&
+		    kbd->accent_table[i].base == ch)
+			return kbd->accent_table[i].result;
+	}
+
+	if (ch == ' ' || ch == d)
+		return d;
+
+	kbd_put_queue(kbd->port, d);
+	return ch;
+}
+
+/*
+ * Handle dead key.
+ */
+static void
+k_dead(struct kbd_data *kbd, unsigned char value)
+{
+	value = ret_diacr[value];
+	kbd->diacr = (kbd->diacr ? handle_diacr(kbd, value) : value);
+}
+
+/*
+ * Normal character handler.
+ */
+static void
+k_self(struct kbd_data *kbd, unsigned char value)
+{
+	if (kbd->diacr)
+		value = handle_diacr(kbd, value);
+	kbd_put_queue(kbd->port, value);
+}
+
+/*
+ * Special key handlers
+ */
+static void
+k_ignore(struct kbd_data *kbd, unsigned char value)
+{
+}
+
+/*
+ * Function key handler.
+ */
+static void
+k_fn(struct kbd_data *kbd, unsigned char value)
+{
+	if (kbd->func_table[value])
+		kbd_puts_queue(kbd->port, kbd->func_table[value]);
+}
+
+static void
+k_spec(struct kbd_data *kbd, unsigned char value)
+{
+	if (value >= NR_FN_HANDLER)
+		return;
+	if (kbd->fn_handler[value])
+		kbd->fn_handler[value](kbd);
+}
+
+/*
+ * Put utf8 character to tty flip buffer.
+ * UTF-8 is defined for words of up to 31 bits,
+ * but we need only 16 bits here
+ */
+static void
+to_utf8(struct tty_port *port, ushort c)
+{
+	if (c < 0x80)
+		/*  0******* */
+		kbd_put_queue(port, c);
+	else if (c < 0x800) {
+		/* 110***** 10****** */
+		kbd_put_queue(port, 0xc0 | (c >> 6));
+		kbd_put_queue(port, 0x80 | (c & 0x3f));
+	} else {
+		/* 1110**** 10****** 10****** */
+		kbd_put_queue(port, 0xe0 | (c >> 12));
+		kbd_put_queue(port, 0x80 | ((c >> 6) & 0x3f));
+		kbd_put_queue(port, 0x80 | (c & 0x3f));
+	}
+}
+
+/*
+ * Process keycode.
+ */
+void
+kbd_keycode(struct kbd_data *kbd, unsigned int keycode)
+{
+	unsigned short keysym;
+	unsigned char type, value;
+
+	if (!kbd)
+		return;
+
+	if (keycode >= 384)
+		keysym = kbd->key_maps[5][keycode - 384];
+	else if (keycode >= 256)
+		keysym = kbd->key_maps[4][keycode - 256];
+	else if (keycode >= 128)
+		keysym = kbd->key_maps[1][keycode - 128];
+	else
+		keysym = kbd->key_maps[0][keycode];
+
+	type = KTYP(keysym);
+	if (type >= 0xf0) {
+		type -= 0xf0;
+		if (type == KT_LETTER)
+			type = KT_LATIN;
+		value = KVAL(keysym);
+#ifdef CONFIG_MAGIC_SYSRQ	       /* Handle the SysRq Hack */
+		if (kbd->sysrq) {
+			if (kbd->sysrq == K(KT_LATIN, '-')) {
+				kbd->sysrq = 0;
+				handle_sysrq(value);
+				return;
+			}
+			if (value == '-') {
+				kbd->sysrq = K(KT_LATIN, '-');
+				return;
+			}
+			/* Incomplete sysrq sequence. */
+			(*k_handler[KTYP(kbd->sysrq)])(kbd, KVAL(kbd->sysrq));
+			kbd->sysrq = 0;
+		} else if ((type == KT_LATIN && value == '^') ||
+			   (type == KT_DEAD && ret_diacr[value] == '^')) {
+			kbd->sysrq = K(type, value);
+			return;
+		}
+#endif
+		(*k_handler[type])(kbd, value);
+	} else
+		to_utf8(kbd->port, keysym);
+}
+
+/*
+ * Ioctl stuff.
+ */
+static int
+do_kdsk_ioctl(struct kbd_data *kbd, struct kbentry __user *user_kbe,
+	      int cmd, int perm)
+{
+	struct kbentry tmp;
+	ushort *key_map, val, ov;
+
+	if (copy_from_user(&tmp, user_kbe, sizeof(struct kbentry)))
+		return -EFAULT;
+#if NR_KEYS < 256
+	if (tmp.kb_index >= NR_KEYS)
+		return -EINVAL;
+#endif
+#if MAX_NR_KEYMAPS < 256
+	if (tmp.kb_table >= MAX_NR_KEYMAPS)
+		return -EINVAL;	
+#endif
+
+	switch (cmd) {
+	case KDGKBENT:
+		key_map = kbd->key_maps[tmp.kb_table];
+		if (key_map) {
+		    val = U(key_map[tmp.kb_index]);
+		    if (KTYP(val) >= KBD_NR_TYPES)
+			val = K_HOLE;
+		} else
+		    val = (tmp.kb_index ? K_HOLE : K_NOSUCHMAP);
+		return put_user(val, &user_kbe->kb_value);
+	case KDSKBENT:
+		if (!perm)
+			return -EPERM;
+		if (!tmp.kb_index && tmp.kb_value == K_NOSUCHMAP) {
+			/* disallocate map */
+			key_map = kbd->key_maps[tmp.kb_table];
+			if (key_map) {
+			    kbd->key_maps[tmp.kb_table] = NULL;
+			    kfree(key_map);
+			}
+			break;
+		}
+
+		if (KTYP(tmp.kb_value) >= KBD_NR_TYPES)
+			return -EINVAL;
+		if (KVAL(tmp.kb_value) > kbd_max_vals[KTYP(tmp.kb_value)])
+			return -EINVAL;
+
+		if (!(key_map = kbd->key_maps[tmp.kb_table])) {
+			int j;
+
+			key_map = kmalloc(sizeof(plain_map),
+						     GFP_KERNEL);
+			if (!key_map)
+				return -ENOMEM;
+			kbd->key_maps[tmp.kb_table] = key_map;
+			for (j = 0; j < NR_KEYS; j++)
+				key_map[j] = U(K_HOLE);
+		}
+		ov = U(key_map[tmp.kb_index]);
+		if (tmp.kb_value == ov)
+			break;	/* nothing to do */
+		/*
+		 * Attention Key.
+		 */
+		if (((ov == K_SAK) || (tmp.kb_value == K_SAK)) &&
+		    !capable(CAP_SYS_ADMIN))
+			return -EPERM;
+		key_map[tmp.kb_index] = U(tmp.kb_value);
+		break;
+	}
+	return 0;
+}
+
+static int
+do_kdgkb_ioctl(struct kbd_data *kbd, struct kbsentry __user *u_kbs,
+	       int cmd, int perm)
+{
+	unsigned char kb_func;
+	char *p;
+	int len;
+
+	/* Get u_kbs->kb_func. */
+	if (get_user(kb_func, &u_kbs->kb_func))
+		return -EFAULT;
+#if MAX_NR_FUNC < 256
+	if (kb_func >= MAX_NR_FUNC)
+		return -EINVAL;
+#endif
+
+	switch (cmd) {
+	case KDGKBSENT:
+		p = kbd->func_table[kb_func];
+		if (p) {
+			len = strlen(p);
+			if (len >= sizeof(u_kbs->kb_string))
+				len = sizeof(u_kbs->kb_string) - 1;
+			if (copy_to_user(u_kbs->kb_string, p, len))
+				return -EFAULT;
+		} else
+			len = 0;
+		if (put_user('\0', u_kbs->kb_string + len))
+			return -EFAULT;
+		break;
+	case KDSKBSENT:
+		if (!perm)
+			return -EPERM;
+		len = strnlen_user(u_kbs->kb_string,
+				   sizeof(u_kbs->kb_string) - 1);
+		if (!len)
+			return -EFAULT;
+		if (len > sizeof(u_kbs->kb_string) - 1)
+			return -EINVAL;
+		p = kmalloc(len + 1, GFP_KERNEL);
+		if (!p)
+			return -ENOMEM;
+		if (copy_from_user(p, u_kbs->kb_string, len)) {
+			kfree(p);
+			return -EFAULT;
+		}
+		p[len] = 0;
+		kfree(kbd->func_table[kb_func]);
+		kbd->func_table[kb_func] = p;
+		break;
+	}
+	return 0;
+}
+
+int kbd_ioctl(struct kbd_data *kbd, unsigned int cmd, unsigned long arg)
+{
+	struct tty_struct *tty;
+	void __user *argp;
+	unsigned int ct;
+	int perm;
+
+	argp = (void __user *)arg;
+
+	/*
+	 * To have permissions to do most of the vt ioctls, we either have
+	 * to be the owner of the tty, or have CAP_SYS_TTY_CONFIG.
+	 */
+	tty = tty_port_tty_get(kbd->port);
+	/* FIXME this test is pretty racy */
+	perm = current->signal->tty == tty || capable(CAP_SYS_TTY_CONFIG);
+	tty_kref_put(tty);
+	switch (cmd) {
+	case KDGKBTYPE:
+		return put_user(KB_101, (char __user *)argp);
+	case KDGKBENT:
+	case KDSKBENT:
+		return do_kdsk_ioctl(kbd, argp, cmd, perm);
+	case KDGKBSENT:
+	case KDSKBSENT:
+		return do_kdgkb_ioctl(kbd, argp, cmd, perm);
+	case KDGKBDIACR:
+	{
+		struct kbdiacrs __user *a = argp;
+		struct kbdiacr diacr;
+		int i;
+
+		if (put_user(kbd->accent_table_size, &a->kb_cnt))
+			return -EFAULT;
+		for (i = 0; i < kbd->accent_table_size; i++) {
+			diacr.diacr = kbd->accent_table[i].diacr;
+			diacr.base = kbd->accent_table[i].base;
+			diacr.result = kbd->accent_table[i].result;
+			if (copy_to_user(a->kbdiacr + i, &diacr, sizeof(struct kbdiacr)))
+			return -EFAULT;
+		}
+		return 0;
+	}
+	case KDGKBDIACRUC:
+	{
+		struct kbdiacrsuc __user *a = argp;
+
+		ct = kbd->accent_table_size;
+		if (put_user(ct, &a->kb_cnt))
+			return -EFAULT;
+		if (copy_to_user(a->kbdiacruc, kbd->accent_table,
+				 ct * sizeof(struct kbdiacruc)))
+			return -EFAULT;
+		return 0;
+	}
+	case KDSKBDIACR:
+	{
+		struct kbdiacrs __user *a = argp;
+		struct kbdiacr diacr;
+		int i;
+
+		if (!perm)
+			return -EPERM;
+		if (get_user(ct, &a->kb_cnt))
+			return -EFAULT;
+		if (ct >= MAX_DIACR)
+			return -EINVAL;
+		kbd->accent_table_size = ct;
+		for (i = 0; i < ct; i++) {
+			if (copy_from_user(&diacr, a->kbdiacr + i, sizeof(struct kbdiacr)))
+				return -EFAULT;
+			kbd->accent_table[i].diacr = diacr.diacr;
+			kbd->accent_table[i].base = diacr.base;
+			kbd->accent_table[i].result = diacr.result;
+		}
+		return 0;
+	}
+	case KDSKBDIACRUC:
+	{
+		struct kbdiacrsuc __user *a = argp;
+
+		if (!perm)
+			return -EPERM;
+		if (get_user(ct, &a->kb_cnt))
+			return -EFAULT;
+		if (ct >= MAX_DIACR)
+			return -EINVAL;
+		kbd->accent_table_size = ct;
+		if (copy_from_user(kbd->accent_table, a->kbdiacruc,
+				   ct * sizeof(struct kbdiacruc)))
+			return -EFAULT;
+		return 0;
+	}
+	default:
+		return -ENOIOCTLCMD;
+	}
+}
+
+EXPORT_SYMBOL(kbd_ioctl);
+EXPORT_SYMBOL(kbd_ascebc);
+EXPORT_SYMBOL(kbd_free);
+EXPORT_SYMBOL(kbd_alloc);
+EXPORT_SYMBOL(kbd_keycode);

drivers/s390/char/keyboard.h

@@ -0,0 +1,56 @@
+/*
+ *    ebcdic keycode functions for s390 console drivers
+ *
+ *    Copyright IBM Corp. 2003
+ *    Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com),
+ */
+
+#include <linux/tty.h>
+#include <linux/tty_flip.h>
+#include <linux/keyboard.h>
+
+#define NR_FN_HANDLER	20
+
+struct kbd_data;
+
+typedef void (fn_handler_fn)(struct kbd_data *);
+
+/*
+ * FIXME: explain key_maps tricks.
+ */
+
+struct kbd_data {
+	struct tty_port *port;
+	unsigned short **key_maps;
+	char **func_table;
+	fn_handler_fn **fn_handler;
+	struct kbdiacruc *accent_table;
+	unsigned int accent_table_size;
+	unsigned int diacr;
+	unsigned short sysrq;
+};
+
+struct kbd_data *kbd_alloc(void);
+void kbd_free(struct kbd_data *);
+void kbd_ascebc(struct kbd_data *, unsigned char *);
+
+void kbd_keycode(struct kbd_data *, unsigned int);
+int kbd_ioctl(struct kbd_data *, unsigned int, unsigned long);
+
+/*
+ * Helper Functions.
+ */
+static inline void
+kbd_put_queue(struct tty_port *port, int ch)
+{
+	tty_insert_flip_char(port, ch, 0);
+	tty_schedule_flip(port);
+}
+
+static inline void
+kbd_puts_queue(struct tty_port *port, char *cp)
+{
+	while (*cp)
+		tty_insert_flip_char(port, *cp++, 0);
+	tty_schedule_flip(port);
+}

drivers/s390/char/monreader.c

@@ -0,0 +1,651 @@
+/*
+ * Character device driver for reading z/VM *MONITOR service records.
+ *
+ * Copyright IBM Corp. 2004, 2009
+ *
+ * Author: Gerald Schaefer <gerald.schaefer@de.ibm.com>
+ */
+
+#define KMSG_COMPONENT "monreader"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/init.h>
+#include <linux/errno.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/miscdevice.h>
+#include <linux/ctype.h>
+#include <linux/spinlock.h>
+#include <linux/interrupt.h>
+#include <linux/poll.h>
+#include <linux/device.h>
+#include <linux/slab.h>
+#include <net/iucv/iucv.h>
+#include <asm/uaccess.h>
+#include <asm/ebcdic.h>
+#include <asm/extmem.h>
+
+
+#define MON_COLLECT_SAMPLE 0x80
+#define MON_COLLECT_EVENT  0x40
+#define MON_SERVICE	   "*MONITOR"
+#define MON_IN_USE	   0x01
+#define MON_MSGLIM	   255
+
+static char mon_dcss_name[9] = "MONDCSS\0";
+
+struct mon_msg {
+	u32 pos;
+	u32 mca_offset;
+	struct iucv_message msg;
+	char msglim_reached;
+	char replied_msglim;
+};
+
+struct mon_private {
+	struct iucv_path *path;
+	struct mon_msg *msg_array[MON_MSGLIM];
+	unsigned int   write_index;
+	unsigned int   read_index;
+	atomic_t msglim_count;
+	atomic_t read_ready;
+	atomic_t iucv_connected;
+	atomic_t iucv_severed;
+};
+
+static unsigned long mon_in_use = 0;
+
+static unsigned long mon_dcss_start;
+static unsigned long mon_dcss_end;
+
+static DECLARE_WAIT_QUEUE_HEAD(mon_read_wait_queue);
+static DECLARE_WAIT_QUEUE_HEAD(mon_conn_wait_queue);
+
+static u8 user_data_connect[16] = {
+	/* Version code, must be 0x01 for shared mode */
+	0x01,
+	/* what to collect */
+	MON_COLLECT_SAMPLE | MON_COLLECT_EVENT,
+	/* DCSS name in EBCDIC, 8 bytes padded with blanks */
+	0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
+	0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
+};
+
+static u8 user_data_sever[16] = {
+	0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
+	0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
+};
+
+static struct device *monreader_device;
+
+/******************************************************************************
+ *                             helper functions                               *
+ *****************************************************************************/
+/*
+ * Create the 8 bytes EBCDIC DCSS segment name from
+ * an ASCII name, incl. padding
+ */
+static void dcss_mkname(char *ascii_name, char *ebcdic_name)
+{
+	int i;
+
+	for (i = 0; i < 8; i++) {
+		if (ascii_name[i] == '\0')
+			break;
+		ebcdic_name[i] = toupper(ascii_name[i]);
+	};
+	for (; i < 8; i++)
+		ebcdic_name[i] = ' ';
+	ASCEBC(ebcdic_name, 8);
+}
+
+static inline unsigned long mon_mca_start(struct mon_msg *monmsg)
+{
+	return *(u32 *) &monmsg->msg.rmmsg;
+}
+
+static inline unsigned long mon_mca_end(struct mon_msg *monmsg)
+{
+	return *(u32 *) &monmsg->msg.rmmsg[4];
+}
+
+static inline u8 mon_mca_type(struct mon_msg *monmsg, u8 index)
+{
+	return *((u8 *) mon_mca_start(monmsg) + monmsg->mca_offset + index);
+}
+
+static inline u32 mon_mca_size(struct mon_msg *monmsg)
+{
+	return mon_mca_end(monmsg) - mon_mca_start(monmsg) + 1;
+}
+
+static inline u32 mon_rec_start(struct mon_msg *monmsg)
+{
+	return *((u32 *) (mon_mca_start(monmsg) + monmsg->mca_offset + 4));
+}
+
+static inline u32 mon_rec_end(struct mon_msg *monmsg)
+{
+	return *((u32 *) (mon_mca_start(monmsg) + monmsg->mca_offset + 8));
+}
+
+static int mon_check_mca(struct mon_msg *monmsg)
+{
+	if ((mon_rec_end(monmsg) <= mon_rec_start(monmsg)) ||
+	    (mon_rec_start(monmsg) < mon_dcss_start) ||
+	    (mon_rec_end(monmsg) > mon_dcss_end) ||
+	    (mon_mca_type(monmsg, 0) == 0) ||
+	    (mon_mca_size(monmsg) % 12 != 0) ||
+	    (mon_mca_end(monmsg) <= mon_mca_start(monmsg)) ||
+	    (mon_mca_end(monmsg) > mon_dcss_end) ||
+	    (mon_mca_start(monmsg) < mon_dcss_start) ||
+	    ((mon_mca_type(monmsg, 1) == 0) && (mon_mca_type(monmsg, 2) == 0)))
+		return -EINVAL;
+	return 0;
+}
+
+static int mon_send_reply(struct mon_msg *monmsg,
+			  struct mon_private *monpriv)
+{
+	int rc;
+
+	rc = iucv_message_reply(monpriv->path, &monmsg->msg,
+				IUCV_IPRMDATA, NULL, 0);
+	atomic_dec(&monpriv->msglim_count);
+	if (likely(!monmsg->msglim_reached)) {
+		monmsg->pos = 0;
+		monmsg->mca_offset = 0;
+		monpriv->read_index = (monpriv->read_index + 1) %
+				      MON_MSGLIM;
+		atomic_dec(&monpriv->read_ready);
+	} else
+		monmsg->replied_msglim = 1;
+	if (rc) {
+		pr_err("Reading monitor data failed with rc=%i\n", rc);
+		return -EIO;
+	}
+	return 0;
+}
+
+static void mon_free_mem(struct mon_private *monpriv)
+{
+	int i;
+
+	for (i = 0; i < MON_MSGLIM; i++)
+		kfree(monpriv->msg_array[i]);
+	kfree(monpriv);
+}
+
+static struct mon_private *mon_alloc_mem(void)
+{
+	int i;
+	struct mon_private *monpriv;
+
+	monpriv = kzalloc(sizeof(struct mon_private), GFP_KERNEL);
+	if (!monpriv)
+		return NULL;
+	for (i = 0; i < MON_MSGLIM; i++) {
+		monpriv->msg_array[i] = kzalloc(sizeof(struct mon_msg),
+						    GFP_KERNEL);
+		if (!monpriv->msg_array[i]) {
+			mon_free_mem(monpriv);
+			return NULL;
+		}
+	}
+	return monpriv;
+}
+
+static inline void mon_next_mca(struct mon_msg *monmsg)
+{
+	if (likely((mon_mca_size(monmsg) - monmsg->mca_offset) == 12))
+		return;
+	monmsg->mca_offset += 12;
+	monmsg->pos = 0;
+}
+
+static struct mon_msg *mon_next_message(struct mon_private *monpriv)
+{
+	struct mon_msg *monmsg;
+
+	if (!atomic_read(&monpriv->read_ready))
+		return NULL;
+	monmsg = monpriv->msg_array[monpriv->read_index];
+	if (unlikely(monmsg->replied_msglim)) {
+		monmsg->replied_msglim = 0;
+		monmsg->msglim_reached = 0;
+		monmsg->pos = 0;
+		monmsg->mca_offset = 0;
+		monpriv->read_index = (monpriv->read_index + 1) %
+				      MON_MSGLIM;
+		atomic_dec(&monpriv->read_ready);
+		return ERR_PTR(-EOVERFLOW);
+	}
+	return monmsg;
+}
+
+
+/******************************************************************************
+ *                               IUCV handler                                 *
+ *****************************************************************************/
+static void mon_iucv_path_complete(struct iucv_path *path, u8 ipuser[16])
+{
+	struct mon_private *monpriv = path->private;
+
+	atomic_set(&monpriv->iucv_connected, 1);
+	wake_up(&mon_conn_wait_queue);
+}
+
+static void mon_iucv_path_severed(struct iucv_path *path, u8 ipuser[16])
+{
+	struct mon_private *monpriv = path->private;
+
+	pr_err("z/VM *MONITOR system service disconnected with rc=%i\n",
+	       ipuser[0]);
+	iucv_path_sever(path, NULL);
+	atomic_set(&monpriv->iucv_severed, 1);
+	wake_up(&mon_conn_wait_queue);
+	wake_up_interruptible(&mon_read_wait_queue);
+}
+
+static void mon_iucv_message_pending(struct iucv_path *path,
+				     struct iucv_message *msg)
+{
+	struct mon_private *monpriv = path->private;
+
+	memcpy(&monpriv->msg_array[monpriv->write_index]->msg,
+	       msg, sizeof(*msg));
+	if (atomic_inc_return(&monpriv->msglim_count) == MON_MSGLIM) {
+		pr_warning("The read queue for monitor data is full\n");
+		monpriv->msg_array[monpriv->write_index]->msglim_reached = 1;
+	}
+	monpriv->write_index = (monpriv->write_index + 1) % MON_MSGLIM;
+	atomic_inc(&monpriv->read_ready);
+	wake_up_interruptible(&mon_read_wait_queue);
+}
+
+static struct iucv_handler monreader_iucv_handler = {
+	.path_complete	 = mon_iucv_path_complete,
+	.path_severed	 = mon_iucv_path_severed,
+	.message_pending = mon_iucv_message_pending,
+};
+
+/******************************************************************************
+ *                               file operations                              *
+ *****************************************************************************/
+static int mon_open(struct inode *inode, struct file *filp)
+{
+	struct mon_private *monpriv;
+	int rc;
+
+	/*
+	 * only one user allowed
+	 */
+	rc = -EBUSY;
+	if (test_and_set_bit(MON_IN_USE, &mon_in_use))
+		goto out;
+
+	rc = -ENOMEM;
+	monpriv = mon_alloc_mem();
+	if (!monpriv)
+		goto out_use;
+
+	/*
+	 * Connect to *MONITOR service
+	 */
+	monpriv->path = iucv_path_alloc(MON_MSGLIM, IUCV_IPRMDATA, GFP_KERNEL);
+	if (!monpriv->path)
+		goto out_priv;
+	rc = iucv_path_connect(monpriv->path, &monreader_iucv_handler,
+			       MON_SERVICE, NULL, user_data_connect, monpriv);
+	if (rc) {
+		pr_err("Connecting to the z/VM *MONITOR system service "
+		       "failed with rc=%i\n", rc);
+		rc = -EIO;
+		goto out_path;
+	}
+	/*
+	 * Wait for connection confirmation
+	 */
+	wait_event(mon_conn_wait_queue,
+		   atomic_read(&monpriv->iucv_connected) ||
+		   atomic_read(&monpriv->iucv_severed));
+	if (atomic_read(&monpriv->iucv_severed)) {
+		atomic_set(&monpriv->iucv_severed, 0);
+		atomic_set(&monpriv->iucv_connected, 0);
+		rc = -EIO;
+		goto out_path;
+	}
+	filp->private_data = monpriv;
+	dev_set_drvdata(monreader_device, monpriv);
+	return nonseekable_open(inode, filp);
+
+out_path:
+	iucv_path_free(monpriv->path);
+out_priv:
+	mon_free_mem(monpriv);
+out_use:
+	clear_bit(MON_IN_USE, &mon_in_use);
+out:
+	return rc;
+}
+
+static int mon_close(struct inode *inode, struct file *filp)
+{
+	int rc, i;
+	struct mon_private *monpriv = filp->private_data;
+
+	/*
+	 * Close IUCV connection and unregister
+	 */
+	if (monpriv->path) {
+		rc = iucv_path_sever(monpriv->path, user_data_sever);
+		if (rc)
+			pr_warning("Disconnecting the z/VM *MONITOR system "
+				   "service failed with rc=%i\n", rc);
+		iucv_path_free(monpriv->path);
+	}
+
+	atomic_set(&monpriv->iucv_severed, 0);
+	atomic_set(&monpriv->iucv_connected, 0);
+	atomic_set(&monpriv->read_ready, 0);
+	atomic_set(&monpriv->msglim_count, 0);
+	monpriv->write_index  = 0;
+	monpriv->read_index   = 0;
+	dev_set_drvdata(monreader_device, NULL);
+
+	for (i = 0; i < MON_MSGLIM; i++)
+		kfree(monpriv->msg_array[i]);
+	kfree(monpriv);
+	clear_bit(MON_IN_USE, &mon_in_use);
+	return 0;
+}
+
+static ssize_t mon_read(struct file *filp, char __user *data,
+			size_t count, loff_t *ppos)
+{
+	struct mon_private *monpriv = filp->private_data;
+	struct mon_msg *monmsg;
+	int ret;
+	u32 mce_start;
+
+	monmsg = mon_next_message(monpriv);
+	if (IS_ERR(monmsg))
+		return PTR_ERR(monmsg);
+
+	if (!monmsg) {
+		if (filp->f_flags & O_NONBLOCK)
+			return -EAGAIN;
+		ret = wait_event_interruptible(mon_read_wait_queue,
+					atomic_read(&monpriv->read_ready) ||
+					atomic_read(&monpriv->iucv_severed));
+		if (ret)
+			return ret;
+		if (unlikely(atomic_read(&monpriv->iucv_severed)))
+			return -EIO;
+		monmsg = monpriv->msg_array[monpriv->read_index];
+	}
+
+	if (!monmsg->pos)
+		monmsg->pos = mon_mca_start(monmsg) + monmsg->mca_offset;
+	if (mon_check_mca(monmsg))
+		goto reply;
+
+	/* read monitor control element (12 bytes) first */
+	mce_start = mon_mca_start(monmsg) + monmsg->mca_offset;
+	if ((monmsg->pos >= mce_start) && (monmsg->pos < mce_start + 12)) {
+		count = min(count, (size_t) mce_start + 12 - monmsg->pos);
+		ret = copy_to_user(data, (void *) (unsigned long) monmsg->pos,
+				   count);
+		if (ret)
+			return -EFAULT;
+		monmsg->pos += count;
+		if (monmsg->pos == mce_start + 12)
+			monmsg->pos = mon_rec_start(monmsg);
+		goto out_copy;
+	}
+
+	/* read records */
+	if (monmsg->pos <= mon_rec_end(monmsg)) {
+		count = min(count, (size_t) mon_rec_end(monmsg) - monmsg->pos
+					    + 1);
+		ret = copy_to_user(data, (void *) (unsigned long) monmsg->pos,
+				   count);
+		if (ret)
+			return -EFAULT;
+		monmsg->pos += count;
+		if (monmsg->pos > mon_rec_end(monmsg))
+			mon_next_mca(monmsg);
+		goto out_copy;
+	}
+reply:
+	ret = mon_send_reply(monmsg, monpriv);
+	return ret;
+
+out_copy:
+	*ppos += count;
+	return count;
+}
+
+static unsigned int mon_poll(struct file *filp, struct poll_table_struct *p)
+{
+	struct mon_private *monpriv = filp->private_data;
+
+	poll_wait(filp, &mon_read_wait_queue, p);
+	if (unlikely(atomic_read(&monpriv->iucv_severed)))
+		return POLLERR;
+	if (atomic_read(&monpriv->read_ready))
+		return POLLIN | POLLRDNORM;
+	return 0;
+}
+
+static const struct file_operations mon_fops = {
+	.owner   = THIS_MODULE,
+	.open    = &mon_open,
+	.release = &mon_close,
+	.read    = &mon_read,
+	.poll    = &mon_poll,
+	.llseek  = noop_llseek,
+};
+
+static struct miscdevice mon_dev = {
+	.name       = "monreader",
+	.fops       = &mon_fops,
+	.minor      = MISC_DYNAMIC_MINOR,
+};
+
+
+/******************************************************************************
+ *				suspend / resume			      *
+ *****************************************************************************/
+static int monreader_freeze(struct device *dev)
+{
+	struct mon_private *monpriv = dev_get_drvdata(dev);
+	int rc;
+
+	if (!monpriv)
+		return 0;
+	if (monpriv->path) {
+		rc = iucv_path_sever(monpriv->path, user_data_sever);
+		if (rc)
+			pr_warning("Disconnecting the z/VM *MONITOR system "
+				   "service failed with rc=%i\n", rc);
+		iucv_path_free(monpriv->path);
+	}
+	atomic_set(&monpriv->iucv_severed, 0);
+	atomic_set(&monpriv->iucv_connected, 0);
+	atomic_set(&monpriv->read_ready, 0);
+	atomic_set(&monpriv->msglim_count, 0);
+	monpriv->write_index  = 0;
+	monpriv->read_index   = 0;
+	monpriv->path = NULL;
+	return 0;
+}
+
+static int monreader_thaw(struct device *dev)
+{
+	struct mon_private *monpriv = dev_get_drvdata(dev);
+	int rc;
+
+	if (!monpriv)
+		return 0;
+	rc = -ENOMEM;
+	monpriv->path = iucv_path_alloc(MON_MSGLIM, IUCV_IPRMDATA, GFP_KERNEL);
+	if (!monpriv->path)
+		goto out;
+	rc = iucv_path_connect(monpriv->path, &monreader_iucv_handler,
+			       MON_SERVICE, NULL, user_data_connect, monpriv);
+	if (rc) {
+		pr_err("Connecting to the z/VM *MONITOR system service "
+		       "failed with rc=%i\n", rc);
+		goto out_path;
+	}
+	wait_event(mon_conn_wait_queue,
+		   atomic_read(&monpriv->iucv_connected) ||
+		   atomic_read(&monpriv->iucv_severed));
+	if (atomic_read(&monpriv->iucv_severed))
+		goto out_path;
+	return 0;
+out_path:
+	rc = -EIO;
+	iucv_path_free(monpriv->path);
+	monpriv->path = NULL;
+out:
+	atomic_set(&monpriv->iucv_severed, 1);
+	return rc;
+}
+
+static int monreader_restore(struct device *dev)
+{
+	int rc;
+
+	segment_unload(mon_dcss_name);
+	rc = segment_load(mon_dcss_name, SEGMENT_SHARED,
+			  &mon_dcss_start, &mon_dcss_end);
+	if (rc < 0) {
+		segment_warning(rc, mon_dcss_name);
+		panic("fatal monreader resume error: no monitor dcss\n");
+	}
+	return monreader_thaw(dev);
+}
+
+static const struct dev_pm_ops monreader_pm_ops = {
+	.freeze  = monreader_freeze,
+	.thaw	 = monreader_thaw,
+	.restore = monreader_restore,
+};
+
+static struct device_driver monreader_driver = {
+	.name = "monreader",
+	.bus  = &iucv_bus,
+	.pm   = &monreader_pm_ops,
+};
+
+
+/******************************************************************************
+ *                              module init/exit                              *
+ *****************************************************************************/
+static int __init mon_init(void)
+{
+	int rc;
+
+	if (!MACHINE_IS_VM) {
+		pr_err("The z/VM *MONITOR record device driver cannot be "
+		       "loaded without z/VM\n");
+		return -ENODEV;
+	}
+
+	/*
+	 * Register with IUCV and connect to *MONITOR service
+	 */
+	rc = iucv_register(&monreader_iucv_handler, 1);
+	if (rc) {
+		pr_err("The z/VM *MONITOR record device driver failed to "
+		       "register with IUCV\n");
+		return rc;
+	}
+
+	rc = driver_register(&monreader_driver);
+	if (rc)
+		goto out_iucv;
+	monreader_device = kzalloc(sizeof(struct device), GFP_KERNEL);
+	if (!monreader_device) {
+		rc = -ENOMEM;
+		goto out_driver;
+	}
+
+	dev_set_name(monreader_device, "monreader-dev");
+	monreader_device->bus = &iucv_bus;
+	monreader_device->parent = iucv_root;
+	monreader_device->driver = &monreader_driver;
+	monreader_device->release = (void (*)(struct device *))kfree;
+	rc = device_register(monreader_device);
+	if (rc) {
+		put_device(monreader_device);
+		goto out_driver;
+	}
+
+	rc = segment_type(mon_dcss_name);
+	if (rc < 0) {
+		segment_warning(rc, mon_dcss_name);
+		goto out_device;
+	}
+	if (rc != SEG_TYPE_SC) {
+		pr_err("The specified *MONITOR DCSS %s does not have the "
+		       "required type SC\n", mon_dcss_name);
+		rc = -EINVAL;
+		goto out_device;
+	}
+
+	rc = segment_load(mon_dcss_name, SEGMENT_SHARED,
+			  &mon_dcss_start, &mon_dcss_end);
+	if (rc < 0) {
+		segment_warning(rc, mon_dcss_name);
+		rc = -EINVAL;
+		goto out_device;
+	}
+	dcss_mkname(mon_dcss_name, &user_data_connect[8]);
+
+	/*
+	 * misc_register() has to be the last action in module_init(), because
+	 * file operations will be available right after this.
+	 */
+	rc = misc_register(&mon_dev);
+	if (rc < 0 )
+		goto out;
+	return 0;
+
+out:
+	segment_unload(mon_dcss_name);
+out_device:
+	device_unregister(monreader_device);
+out_driver:
+	driver_unregister(&monreader_driver);
+out_iucv:
+	iucv_unregister(&monreader_iucv_handler, 1);
+	return rc;
+}
+
+static void __exit mon_exit(void)
+{
+	segment_unload(mon_dcss_name);
+	misc_deregister(&mon_dev);
+	device_unregister(monreader_device);
+	driver_unregister(&monreader_driver);
+	iucv_unregister(&monreader_iucv_handler, 1);
+	return;
+}
+
+
+module_init(mon_init);
+module_exit(mon_exit);
+
+module_param_string(mondcss, mon_dcss_name, 9, 0444);
+MODULE_PARM_DESC(mondcss, "Name of DCSS segment to be used for *MONITOR "
+		 "service, max. 8 chars. Default is MONDCSS");
+
+MODULE_AUTHOR("Gerald Schaefer <geraldsc@de.ibm.com>");
+MODULE_DESCRIPTION("Character device driver for reading z/VM "
+		   "monitor service records.");
+MODULE_LICENSE("GPL");

drivers/s390/char/monwriter.c

@@ -0,0 +1,399 @@
+/*
+ * Character device driver for writing z/VM *MONITOR service records.
+ *
+ * Copyright IBM Corp. 2006, 2009
+ *
+ * Author(s): Melissa Howland <Melissa.Howland@us.ibm.com>
+ */
+
+#define KMSG_COMPONENT "monwriter"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/init.h>
+#include <linux/errno.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/miscdevice.h>
+#include <linux/ctype.h>
+#include <linux/poll.h>
+#include <linux/mutex.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+#include <asm/uaccess.h>
+#include <asm/ebcdic.h>
+#include <asm/io.h>
+#include <asm/appldata.h>
+#include <asm/monwriter.h>
+
+#define MONWRITE_MAX_DATALEN	4010
+
+static int mon_max_bufs = 255;
+static int mon_buf_count;
+
+struct mon_buf {
+	struct list_head list;
+	struct monwrite_hdr hdr;
+	int diag_done;
+	char *data;
+};
+
+static LIST_HEAD(mon_priv_list);
+
+struct mon_private {
+	struct list_head priv_list;
+	struct list_head list;
+	struct monwrite_hdr hdr;
+	size_t hdr_to_read;
+	size_t data_to_read;
+	struct mon_buf *current_buf;
+	struct mutex thread_mutex;
+};
+
+/*
+ * helper functions
+ */
+
+static int monwrite_diag(struct monwrite_hdr *myhdr, char *buffer, int fcn)
+{
+	struct appldata_product_id id;
+	int rc;
+
+	strncpy(id.prod_nr, "LNXAPPL", 7);
+	id.prod_fn = myhdr->applid;
+	id.record_nr = myhdr->record_num;
+	id.version_nr = myhdr->version;
+	id.release_nr = myhdr->release;
+	id.mod_lvl = myhdr->mod_level;
+	rc = appldata_asm(&id, fcn, (void *) buffer, myhdr->datalen);
+	if (rc <= 0)
+		return rc;
+	pr_err("Writing monitor data failed with rc=%i\n", rc);
+	if (rc == 5)
+		return -EPERM;
+	return -EINVAL;
+}
+
+static struct mon_buf *monwrite_find_hdr(struct mon_private *monpriv,
+					 struct monwrite_hdr *monhdr)
+{
+	struct mon_buf *entry, *next;
+
+	list_for_each_entry_safe(entry, next, &monpriv->list, list)
+		if ((entry->hdr.mon_function == monhdr->mon_function ||
+		     monhdr->mon_function == MONWRITE_STOP_INTERVAL) &&
+		    entry->hdr.applid == monhdr->applid &&
+		    entry->hdr.record_num == monhdr->record_num &&
+		    entry->hdr.version == monhdr->version &&
+		    entry->hdr.release == monhdr->release &&
+		    entry->hdr.mod_level == monhdr->mod_level)
+			return entry;
+
+	return NULL;
+}
+
+static int monwrite_new_hdr(struct mon_private *monpriv)
+{
+	struct monwrite_hdr *monhdr = &monpriv->hdr;
+	struct mon_buf *monbuf;
+	int rc = 0;
+
+	if (monhdr->datalen > MONWRITE_MAX_DATALEN ||
+	    monhdr->mon_function > MONWRITE_START_CONFIG ||
+	    monhdr->hdrlen != sizeof(struct monwrite_hdr))
+		return -EINVAL;
+	monbuf = NULL;
+	if (monhdr->mon_function != MONWRITE_GEN_EVENT)
+		monbuf = monwrite_find_hdr(monpriv, monhdr);
+	if (monbuf) {
+		if (monhdr->mon_function == MONWRITE_STOP_INTERVAL) {
+			monhdr->datalen = monbuf->hdr.datalen;
+			rc = monwrite_diag(monhdr, monbuf->data,
+					   APPLDATA_STOP_REC);
+			list_del(&monbuf->list);
+			mon_buf_count--;
+			kfree(monbuf->data);
+			kfree(monbuf);
+			monbuf = NULL;
+		}
+	} else if (monhdr->mon_function != MONWRITE_STOP_INTERVAL) {
+		if (mon_buf_count >= mon_max_bufs)
+			return -ENOSPC;
+		monbuf = kzalloc(sizeof(struct mon_buf), GFP_KERNEL);
+		if (!monbuf)
+			return -ENOMEM;
+		monbuf->data = kzalloc(monhdr->datalen,
+				       GFP_KERNEL | GFP_DMA);
+		if (!monbuf->data) {
+			kfree(monbuf);
+			return -ENOMEM;
+		}
+		monbuf->hdr = *monhdr;
+		list_add_tail(&monbuf->list, &monpriv->list);
+		if (monhdr->mon_function != MONWRITE_GEN_EVENT)
+			mon_buf_count++;
+	}
+	monpriv->current_buf = monbuf;
+	return rc;
+}
+
+static int monwrite_new_data(struct mon_private *monpriv)
+{
+	struct monwrite_hdr *monhdr = &monpriv->hdr;
+	struct mon_buf *monbuf = monpriv->current_buf;
+	int rc = 0;
+
+	switch (monhdr->mon_function) {
+	case MONWRITE_START_INTERVAL:
+		if (!monbuf->diag_done) {
+			rc = monwrite_diag(monhdr, monbuf->data,
+					   APPLDATA_START_INTERVAL_REC);
+			monbuf->diag_done = 1;
+		}
+		break;
+	case MONWRITE_START_CONFIG:
+		if (!monbuf->diag_done) {
+			rc = monwrite_diag(monhdr, monbuf->data,
+					   APPLDATA_START_CONFIG_REC);
+			monbuf->diag_done = 1;
+		}
+		break;
+	case MONWRITE_GEN_EVENT:
+		rc = monwrite_diag(monhdr, monbuf->data,
+				   APPLDATA_GEN_EVENT_REC);
+		list_del(&monpriv->current_buf->list);
+		kfree(monpriv->current_buf->data);
+		kfree(monpriv->current_buf);
+		monpriv->current_buf = NULL;
+		break;
+	default:
+		/* monhdr->mon_function is checked in monwrite_new_hdr */
+		BUG();
+	}
+	return rc;
+}
+
+/*
+ * file operations
+ */
+
+static int monwrite_open(struct inode *inode, struct file *filp)
+{
+	struct mon_private *monpriv;
+
+	monpriv = kzalloc(sizeof(struct mon_private), GFP_KERNEL);
+	if (!monpriv)
+		return -ENOMEM;
+	INIT_LIST_HEAD(&monpriv->list);
+	monpriv->hdr_to_read = sizeof(monpriv->hdr);
+	mutex_init(&monpriv->thread_mutex);
+	filp->private_data = monpriv;
+	list_add_tail(&monpriv->priv_list, &mon_priv_list);
+	return nonseekable_open(inode, filp);
+}
+
+static int monwrite_close(struct inode *inode, struct file *filp)
+{
+	struct mon_private *monpriv = filp->private_data;
+	struct mon_buf *entry, *next;
+
+	list_for_each_entry_safe(entry, next, &monpriv->list, list) {
+		if (entry->hdr.mon_function != MONWRITE_GEN_EVENT)
+			monwrite_diag(&entry->hdr, entry->data,
+				      APPLDATA_STOP_REC);
+		mon_buf_count--;
+		list_del(&entry->list);
+		kfree(entry->data);
+		kfree(entry);
+	}
+	list_del(&monpriv->priv_list);
+	kfree(monpriv);
+	return 0;
+}
+
+static ssize_t monwrite_write(struct file *filp, const char __user *data,
+			      size_t count, loff_t *ppos)
+{
+	struct mon_private *monpriv = filp->private_data;
+	size_t len, written;
+	void *to;
+	int rc;
+
+	mutex_lock(&monpriv->thread_mutex);
+	for (written = 0; written < count; ) {
+		if (monpriv->hdr_to_read) {
+			len = min(count - written, monpriv->hdr_to_read);
+			to = (char *) &monpriv->hdr +
+				sizeof(monpriv->hdr) - monpriv->hdr_to_read;
+			if (copy_from_user(to, data + written, len)) {
+				rc = -EFAULT;
+				goto out_error;
+			}
+			monpriv->hdr_to_read -= len;
+			written += len;
+			if (monpriv->hdr_to_read > 0)
+				continue;
+			rc = monwrite_new_hdr(monpriv);
+			if (rc)
+				goto out_error;
+			monpriv->data_to_read = monpriv->current_buf ?
+				monpriv->current_buf->hdr.datalen : 0;
+		}
+
+		if (monpriv->data_to_read) {
+			len = min(count - written, monpriv->data_to_read);
+			to = monpriv->current_buf->data +
+				monpriv->hdr.datalen - monpriv->data_to_read;
+			if (copy_from_user(to, data + written, len)) {
+				rc = -EFAULT;
+				goto out_error;
+			}
+			monpriv->data_to_read -= len;
+			written += len;
+			if (monpriv->data_to_read > 0)
+				continue;
+			rc = monwrite_new_data(monpriv);
+			if (rc)
+				goto out_error;
+		}
+		monpriv->hdr_to_read = sizeof(monpriv->hdr);
+	}
+	mutex_unlock(&monpriv->thread_mutex);
+	return written;
+
+out_error:
+	monpriv->data_to_read = 0;
+	monpriv->hdr_to_read = sizeof(struct monwrite_hdr);
+	mutex_unlock(&monpriv->thread_mutex);
+	return rc;
+}
+
+static const struct file_operations monwrite_fops = {
+	.owner	 = THIS_MODULE,
+	.open	 = &monwrite_open,
+	.release = &monwrite_close,
+	.write	 = &monwrite_write,
+	.llseek  = noop_llseek,
+};
+
+static struct miscdevice mon_dev = {
+	.name	= "monwriter",
+	.fops	= &monwrite_fops,
+	.minor	= MISC_DYNAMIC_MINOR,
+};
+
+/*
+ * suspend/resume
+ */
+
+static int monwriter_freeze(struct device *dev)
+{
+	struct mon_private *monpriv;
+	struct mon_buf *monbuf;
+
+	list_for_each_entry(monpriv, &mon_priv_list, priv_list) {
+		list_for_each_entry(monbuf, &monpriv->list, list) {
+			if (monbuf->hdr.mon_function != MONWRITE_GEN_EVENT)
+				monwrite_diag(&monbuf->hdr, monbuf->data,
+					      APPLDATA_STOP_REC);
+		}
+	}
+	return 0;
+}
+
+static int monwriter_restore(struct device *dev)
+{
+	struct mon_private *monpriv;
+	struct mon_buf *monbuf;
+
+	list_for_each_entry(monpriv, &mon_priv_list, priv_list) {
+		list_for_each_entry(monbuf, &monpriv->list, list) {
+			if (monbuf->hdr.mon_function == MONWRITE_START_INTERVAL)
+				monwrite_diag(&monbuf->hdr, monbuf->data,
+					      APPLDATA_START_INTERVAL_REC);
+			if (monbuf->hdr.mon_function == MONWRITE_START_CONFIG)
+				monwrite_diag(&monbuf->hdr, monbuf->data,
+					      APPLDATA_START_CONFIG_REC);
+		}
+	}
+	return 0;
+}
+
+static int monwriter_thaw(struct device *dev)
+{
+	return monwriter_restore(dev);
+}
+
+static const struct dev_pm_ops monwriter_pm_ops = {
+	.freeze		= monwriter_freeze,
+	.thaw		= monwriter_thaw,
+	.restore	= monwriter_restore,
+};
+
+static struct platform_driver monwriter_pdrv = {
+	.driver = {
+		.name	= "monwriter",
+		.owner	= THIS_MODULE,
+		.pm	= &monwriter_pm_ops,
+	},
+};
+
+static struct platform_device *monwriter_pdev;
+
+/*
+ * module init/exit
+ */
+
+static int __init mon_init(void)
+{
+	int rc;
+
+	if (!MACHINE_IS_VM)
+		return -ENODEV;
+
+	rc = platform_driver_register(&monwriter_pdrv);
+	if (rc)
+		return rc;
+
+	monwriter_pdev = platform_device_register_simple("monwriter", -1, NULL,
+							0);
+	if (IS_ERR(monwriter_pdev)) {
+		rc = PTR_ERR(monwriter_pdev);
+		goto out_driver;
+	}
+
+	/*
+	 * misc_register() has to be the last action in module_init(), because
+	 * file operations will be available right after this.
+	 */
+	rc = misc_register(&mon_dev);
+	if (rc)
+		goto out_device;
+	return 0;
+
+out_device:
+	platform_device_unregister(monwriter_pdev);
+out_driver:
+	platform_driver_unregister(&monwriter_pdrv);
+	return rc;
+}
+
+static void __exit mon_exit(void)
+{
+	misc_deregister(&mon_dev);
+	platform_device_unregister(monwriter_pdev);
+	platform_driver_unregister(&monwriter_pdrv);
+}
+
+module_init(mon_init);
+module_exit(mon_exit);
+
+module_param_named(max_bufs, mon_max_bufs, int, 0644);
+MODULE_PARM_DESC(max_bufs, "Maximum number of sample monitor data buffers "
+		 "that can be active at one time");
+
+MODULE_AUTHOR("Melissa Howland <Melissa.Howland@us.ibm.com>");
+MODULE_DESCRIPTION("Character device driver for writing z/VM "
+		   "APPLDATA monitor records.");
+MODULE_LICENSE("GPL");

drivers/s390/char/raw3270.h

@@ -0,0 +1,288 @@
+/*
+ * IBM/3270 Driver
+ *
+ * Author(s):
+ *   Original 3270 Code for 2.4 written by Richard Hitt (UTS Global)
+ *   Rewritten for 2.5 by Martin Schwidefsky <schwidefsky@de.ibm.com>
+ *     Copyright IBM Corp. 2003, 2009
+ */
+
+#include <asm/idals.h>
+#include <asm/ioctl.h>
+
+/* ioctls for fullscreen 3270 */
+#define TUBICMD		_IO('3', 3)	/* set ccw command for fs reads. */
+#define TUBOCMD		_IO('3', 4)	/* set ccw command for fs writes. */
+#define TUBGETI		_IO('3', 7)	/* get ccw command for fs reads. */
+#define TUBGETO		_IO('3', 8)	/* get ccw command for fs writes. */
+#define TUBSETMOD	_IO('3',12)	/* FIXME: what does it do ?*/
+#define TUBGETMOD	_IO('3',13)	/* FIXME: what does it do ?*/
+
+/* Local Channel Commands */
+#define TC_WRITE	0x01		/* Write */
+#define TC_RDBUF	0x02		/* Read Buffer */
+#define TC_EWRITE	0x05		/* Erase write */
+#define TC_READMOD	0x06		/* Read modified */
+#define TC_EWRITEA	0x0d		/* Erase write alternate */
+#define TC_WRITESF	0x11		/* Write structured field */
+
+/* Buffer Control Orders */
+#define TO_SF		0x1d		/* Start field */
+#define TO_SBA		0x11		/* Set buffer address */
+#define TO_IC		0x13		/* Insert cursor */
+#define TO_PT		0x05		/* Program tab */
+#define TO_RA		0x3c		/* Repeat to address */
+#define TO_SFE		0x29		/* Start field extended */
+#define TO_EUA		0x12		/* Erase unprotected to address */
+#define TO_MF		0x2c		/* Modify field */
+#define TO_SA		0x28		/* Set attribute */
+
+/* Field Attribute Bytes */
+#define TF_INPUT	0x40		/* Visible input */
+#define TF_INPUTN	0x4c		/* Invisible input */
+#define TF_INMDT	0xc1		/* Visible, Set-MDT */
+#define TF_LOG		0x60
+
+/* Character Attribute Bytes */
+#define TAT_RESET	0x00
+#define TAT_FIELD	0xc0
+#define TAT_EXTHI	0x41
+#define TAT_COLOR	0x42
+#define TAT_CHARS	0x43
+#define TAT_TRANS	0x46
+
+/* Extended-Highlighting Bytes */
+#define TAX_RESET	0x00
+#define TAX_BLINK	0xf1
+#define TAX_REVER	0xf2
+#define TAX_UNDER	0xf4
+
+/* Reset value */
+#define TAR_RESET	0x00
+
+/* Color values */
+#define TAC_RESET	0x00
+#define TAC_BLUE	0xf1
+#define TAC_RED		0xf2
+#define TAC_PINK	0xf3
+#define TAC_GREEN	0xf4
+#define TAC_TURQ	0xf5
+#define TAC_YELLOW	0xf6
+#define TAC_WHITE	0xf7
+#define TAC_DEFAULT	0x00
+
+/* Write Control Characters */
+#define TW_NONE		0x40		/* No particular action */
+#define TW_KR		0xc2		/* Keyboard restore */
+#define TW_PLUSALARM	0x04		/* Add this bit for alarm */
+
+#define RAW3270_FIRSTMINOR	1	/* First minor number */
+#define RAW3270_MAXDEVS		255	/* Max number of 3270 devices */
+
+/* For TUBGETMOD and TUBSETMOD. Should include. */
+struct raw3270_iocb {
+	short model;
+	short line_cnt;
+	short col_cnt;
+	short pf_cnt;
+	short re_cnt;
+	short map;
+};
+
+struct raw3270;
+struct raw3270_view;
+extern struct class *class3270;
+
+/* 3270 CCW request */
+struct raw3270_request {
+	struct list_head list;		/* list head for request queueing. */
+	struct raw3270_view *view;	/* view of this request */
+	struct ccw1 ccw;		/* single ccw. */
+	void *buffer;			/* output buffer. */
+	size_t size;			/* size of output buffer. */
+	int rescnt;			/* residual count from devstat. */
+	int rc;				/* return code for this request. */
+
+	/* Callback for delivering final status. */
+	void (*callback)(struct raw3270_request *, void *);
+	void *callback_data;
+};
+
+struct raw3270_request *raw3270_request_alloc(size_t size);
+struct raw3270_request *raw3270_request_alloc_bootmem(size_t size);
+void raw3270_request_free(struct raw3270_request *);
+void raw3270_request_reset(struct raw3270_request *);
+void raw3270_request_set_cmd(struct raw3270_request *, u8 cmd);
+int  raw3270_request_add_data(struct raw3270_request *, void *, size_t);
+void raw3270_request_set_data(struct raw3270_request *, void *, size_t);
+void raw3270_request_set_idal(struct raw3270_request *, struct idal_buffer *);
+
+static inline int
+raw3270_request_final(struct raw3270_request *rq)
+{
+	return list_empty(&rq->list);
+}
+
+void raw3270_buffer_address(struct raw3270 *, char *, unsigned short);
+
+/* Return value of *intv (see raw3270_fn below) can be one of the following: */
+#define RAW3270_IO_DONE		0	/* request finished */
+#define RAW3270_IO_BUSY		1	/* request still active */
+#define RAW3270_IO_RETRY	2	/* retry current request */
+#define RAW3270_IO_STOP		3	/* kill current request */
+
+/*
+ * Functions of a 3270 view.
+ */
+struct raw3270_fn {
+	int  (*activate)(struct raw3270_view *);
+	void (*deactivate)(struct raw3270_view *);
+	int  (*intv)(struct raw3270_view *,
+		     struct raw3270_request *, struct irb *);
+	void (*release)(struct raw3270_view *);
+	void (*free)(struct raw3270_view *);
+	void (*resize)(struct raw3270_view *, int, int, int);
+};
+
+/*
+ * View structure chaining. The raw3270_view structure is meant to
+ * be embedded at the start of the real view data structure, e.g.:
+ *   struct example {
+ *     struct raw3270_view view;
+ *     ...
+ *   };
+ */
+struct raw3270_view {
+	struct list_head list;
+	spinlock_t lock;
+	atomic_t ref_count;
+	struct raw3270 *dev;
+	struct raw3270_fn *fn;
+	unsigned int model;
+	unsigned int rows, cols;	/* # of rows & colums of the view */
+	unsigned char *ascebc;		/* ascii -> ebcdic table */
+};
+
+int raw3270_add_view(struct raw3270_view *, struct raw3270_fn *, int);
+int raw3270_activate_view(struct raw3270_view *);
+void raw3270_del_view(struct raw3270_view *);
+void raw3270_deactivate_view(struct raw3270_view *);
+struct raw3270_view *raw3270_find_view(struct raw3270_fn *, int);
+int raw3270_start(struct raw3270_view *, struct raw3270_request *);
+int raw3270_start_locked(struct raw3270_view *, struct raw3270_request *);
+int raw3270_start_irq(struct raw3270_view *, struct raw3270_request *);
+int raw3270_reset(struct raw3270_view *);
+struct raw3270_view *raw3270_view(struct raw3270_view *);
+int raw3270_view_active(struct raw3270_view *);
+
+/* Reference count inliner for view structures. */
+static inline void
+raw3270_get_view(struct raw3270_view *view)
+{
+	atomic_inc(&view->ref_count);
+}
+
+extern wait_queue_head_t raw3270_wait_queue;
+
+static inline void
+raw3270_put_view(struct raw3270_view *view)
+{
+	if (atomic_dec_return(&view->ref_count) == 0)
+		wake_up(&raw3270_wait_queue);
+}
+
+struct raw3270 *raw3270_setup_console(void);
+void raw3270_wait_cons_dev(struct raw3270 *);
+
+/* Notifier for device addition/removal */
+struct raw3270_notifier {
+	struct list_head list;
+	void (*create)(int minor);
+	void (*destroy)(int minor);
+};
+
+int raw3270_register_notifier(struct raw3270_notifier *);
+void raw3270_unregister_notifier(struct raw3270_notifier *);
+void raw3270_pm_unfreeze(struct raw3270_view *);
+
+/*
+ * Little memory allocator for string objects. 
+ */
+struct string
+{
+	struct list_head list;
+	struct list_head update;
+	unsigned long size;
+	unsigned long len;
+	char string[0];
+} __attribute__ ((aligned(8)));
+
+static inline struct string *
+alloc_string(struct list_head *free_list, unsigned long len)
+{
+	struct string *cs, *tmp;
+	unsigned long size;
+
+	size = (len + 7L) & -8L;
+	list_for_each_entry(cs, free_list, list) {
+		if (cs->size < size)
+			continue;
+		if (cs->size > size + sizeof(struct string)) {
+			char *endaddr = (char *) (cs + 1) + cs->size;
+			tmp = (struct string *) (endaddr - size) - 1;
+			tmp->size = size;
+			cs->size -= size + sizeof(struct string);
+			cs = tmp;
+		} else
+			list_del(&cs->list);
+		cs->len = len;
+		INIT_LIST_HEAD(&cs->list);
+		INIT_LIST_HEAD(&cs->update);
+		return cs;
+	}
+	return NULL;
+}
+
+static inline unsigned long
+free_string(struct list_head *free_list, struct string *cs)
+{
+	struct string *tmp;
+	struct list_head *p, *left;
+
+	/* Find out the left neighbour in free memory list. */
+	left = free_list;
+	list_for_each(p, free_list) {
+		if (list_entry(p, struct string, list) > cs)
+			break;
+		left = p;
+	}
+	/* Try to merge with right neighbour = next element from left. */
+	if (left->next != free_list) {
+		tmp = list_entry(left->next, struct string, list);
+		if ((char *) (cs + 1) + cs->size == (char *) tmp) {
+			list_del(&tmp->list);
+			cs->size += tmp->size + sizeof(struct string);
+		}
+	}
+	/* Try to merge with left neighbour. */
+	if (left != free_list) {
+		tmp = list_entry(left, struct string, list);
+		if ((char *) (tmp + 1) + tmp->size == (char *) cs) {
+			tmp->size += cs->size + sizeof(struct string);
+			return tmp->size;
+		}
+	}
+	__list_add(&cs->list, left, left->next);
+	return cs->size;
+}
+
+static inline void
+add_string_memory(struct list_head *free_list, void *mem, unsigned long size)
+{
+	struct string *cs;
+
+	cs = (struct string *) mem;
+	cs->size = size - sizeof(struct string);
+	free_string(free_list, cs);
+}
+

drivers/s390/char/sclp.h

@@ -0,0 +1,244 @@
+/*
+ * Copyright IBM Corp. 1999,2012
+ *
+ * Author(s): Martin Peschke <mpeschke@de.ibm.com>
+ *	      Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#ifndef __SCLP_H__
+#define __SCLP_H__
+
+#include <linux/types.h>
+#include <linux/list.h>
+#include <asm/sclp.h>
+#include <asm/ebcdic.h>
+
+/* maximum number of pages concerning our own memory management */
+#define MAX_KMEM_PAGES (sizeof(unsigned long) << 3)
+#define SCLP_CONSOLE_PAGES	6
+
+#define EVTYP_OPCMD		0x01
+#define EVTYP_MSG		0x02
+#define EVTYP_DIAG_TEST		0x07
+#define EVTYP_STATECHANGE	0x08
+#define EVTYP_PMSGCMD		0x09
+#define EVTYP_CNTLPROGOPCMD	0x20
+#define EVTYP_CNTLPROGIDENT	0x0B
+#define EVTYP_SIGQUIESCE	0x1D
+#define EVTYP_VT220MSG		0x1A
+#define EVTYP_CONFMGMDATA	0x04
+#define EVTYP_SDIAS		0x1C
+#define EVTYP_ASYNC		0x0A
+#define EVTYP_OCF		0x1E
+
+#define EVTYP_OPCMD_MASK	0x80000000
+#define EVTYP_MSG_MASK		0x40000000
+#define EVTYP_DIAG_TEST_MASK	0x02000000
+#define EVTYP_STATECHANGE_MASK	0x01000000
+#define EVTYP_PMSGCMD_MASK	0x00800000
+#define EVTYP_CTLPROGOPCMD_MASK	0x00000001
+#define EVTYP_CTLPROGIDENT_MASK	0x00200000
+#define EVTYP_SIGQUIESCE_MASK	0x00000008
+#define EVTYP_VT220MSG_MASK	0x00000040
+#define EVTYP_CONFMGMDATA_MASK	0x10000000
+#define EVTYP_SDIAS_MASK	0x00000010
+#define EVTYP_ASYNC_MASK	0x00400000
+#define EVTYP_OCF_MASK		0x00000004
+
+#define GNRLMSGFLGS_DOM		0x8000
+#define GNRLMSGFLGS_SNDALRM	0x4000
+#define GNRLMSGFLGS_HOLDMSG	0x2000
+
+#define LNTPFLGS_CNTLTEXT	0x8000
+#define LNTPFLGS_LABELTEXT	0x4000
+#define LNTPFLGS_DATATEXT	0x2000
+#define LNTPFLGS_ENDTEXT	0x1000
+#define LNTPFLGS_PROMPTTEXT	0x0800
+
+typedef unsigned int sclp_cmdw_t;
+
+#define SCLP_CMDW_READ_EVENT_DATA	0x00770005
+#define SCLP_CMDW_WRITE_EVENT_DATA	0x00760005
+#define SCLP_CMDW_WRITE_EVENT_MASK	0x00780005
+
+#define GDS_ID_MDSMU		0x1310
+#define GDS_ID_MDSROUTEINFO	0x1311
+#define GDS_ID_AGUNWRKCORR	0x1549
+#define GDS_ID_SNACONDREPORT	0x1532
+#define GDS_ID_CPMSU		0x1212
+#define GDS_ID_ROUTTARGINSTR	0x154D
+#define GDS_ID_OPREQ		0x8070
+#define GDS_ID_TEXTCMD		0x1320
+
+#define GDS_KEY_SELFDEFTEXTMSG	0x31
+
+enum sclp_pm_event {
+	SCLP_PM_EVENT_FREEZE,
+	SCLP_PM_EVENT_THAW,
+	SCLP_PM_EVENT_RESTORE,
+};
+
+#define SCLP_PANIC_PRIO		1
+#define SCLP_PANIC_PRIO_CLIENT	0
+
+typedef u32 sccb_mask_t;	/* ATTENTION: assumes 32bit mask !!! */
+
+struct sccb_header {
+	u16	length;
+	u8	function_code;
+	u8	control_mask[3];
+	u16	response_code;
+} __attribute__((packed));
+
+struct init_sccb {
+	struct sccb_header header;
+	u16 _reserved;
+	u16 mask_length;
+	sccb_mask_t receive_mask;
+	sccb_mask_t send_mask;
+	sccb_mask_t sclp_receive_mask;
+	sccb_mask_t sclp_send_mask;
+} __attribute__((packed));
+
+extern u64 sclp_facilities;
+
+#define SCLP_HAS_CHP_INFO	(sclp_facilities & 0x8000000000000000ULL)
+#define SCLP_HAS_CHP_RECONFIG	(sclp_facilities & 0x2000000000000000ULL)
+#define SCLP_HAS_CPU_INFO	(sclp_facilities & 0x0800000000000000ULL)
+#define SCLP_HAS_CPU_RECONFIG	(sclp_facilities & 0x0400000000000000ULL)
+#define SCLP_HAS_PCI_RECONFIG	(sclp_facilities & 0x0000000040000000ULL)
+
+
+struct gds_subvector {
+	u8	length;
+	u8	key;
+} __attribute__((packed));
+
+struct gds_vector {
+	u16	length;
+	u16	gds_id;
+} __attribute__((packed));
+
+struct evbuf_header {
+	u16	length;
+	u8	type;
+	u8	flags;
+	u16	_reserved;
+} __attribute__((packed));
+
+struct sclp_req {
+	struct list_head list;		/* list_head for request queueing. */
+	sclp_cmdw_t command;		/* sclp command to execute */
+	void	*sccb;			/* pointer to the sccb to execute */
+	char	status;			/* status of this request */
+	int     start_count;		/* number of SVCs done for this req */
+	/* Callback that is called after reaching final status. */
+	void (*callback)(struct sclp_req *, void *data);
+	void *callback_data;
+	int queue_timeout;		/* request queue timeout (sec), set by
+					   caller of sclp_add_request(), if
+					   needed */
+	/* Internal fields */
+	unsigned long queue_expires;	/* request queue timeout (jiffies) */
+};
+
+#define SCLP_REQ_FILLED	  0x00	/* request is ready to be processed */
+#define SCLP_REQ_QUEUED	  0x01	/* request is queued to be processed */
+#define SCLP_REQ_RUNNING  0x02	/* request is currently running */
+#define SCLP_REQ_DONE	  0x03	/* request is completed successfully */
+#define SCLP_REQ_FAILED	  0x05	/* request is finally failed */
+#define SCLP_REQ_QUEUED_TIMEOUT 0x06	/* request on queue timed out */
+
+#define SCLP_QUEUE_INTERVAL 5	/* timeout interval for request queue */
+
+/* function pointers that a high level driver has to use for registration */
+/* of some routines it wants to be called from the low level driver */
+struct sclp_register {
+	struct list_head list;
+	/* User wants to receive: */
+	sccb_mask_t receive_mask;
+	/* User wants to send: */
+	sccb_mask_t send_mask;
+	/* H/W can receive: */
+	sccb_mask_t sclp_receive_mask;
+	/* H/W can send: */
+	sccb_mask_t sclp_send_mask;
+	/* called if event type availability changes */
+	void (*state_change_fn)(struct sclp_register *);
+	/* called for events in cp_receive_mask/sclp_receive_mask */
+	void (*receiver_fn)(struct evbuf_header *);
+	/* called for power management events */
+	void (*pm_event_fn)(struct sclp_register *, enum sclp_pm_event);
+	/* pm event posted flag */
+	int pm_event_posted;
+};
+
+/* externals from sclp.c */
+int sclp_add_request(struct sclp_req *req);
+void sclp_sync_wait(void);
+int sclp_register(struct sclp_register *reg);
+void sclp_unregister(struct sclp_register *reg);
+int sclp_remove_processed(struct sccb_header *sccb);
+int sclp_deactivate(void);
+int sclp_reactivate(void);
+int sclp_service_call(sclp_cmdw_t command, void *sccb);
+int sclp_sync_request(sclp_cmdw_t command, void *sccb);
+int sclp_sync_request_timeout(sclp_cmdw_t command, void *sccb, int timeout);
+
+int sclp_sdias_init(void);
+void sclp_sdias_exit(void);
+
+extern int sclp_console_pages;
+extern int sclp_console_drop;
+extern unsigned long sclp_console_full;
+extern u8 sclp_fac84;
+extern unsigned long long sclp_rzm;
+extern unsigned long long sclp_rnmax;
+
+/* useful inlines */
+
+/* VM uses EBCDIC 037, LPAR+native(SE+HMC) use EBCDIC 500 */
+/* translate single character from ASCII to EBCDIC */
+static inline unsigned char
+sclp_ascebc(unsigned char ch)
+{
+	return (MACHINE_IS_VM) ? _ascebc[ch] : _ascebc_500[ch];
+}
+
+/* translate string from EBCDIC to ASCII */
+static inline void
+sclp_ebcasc_str(unsigned char *str, int nr)
+{
+	(MACHINE_IS_VM) ? EBCASC(str, nr) : EBCASC_500(str, nr);
+}
+
+/* translate string from ASCII to EBCDIC */
+static inline void
+sclp_ascebc_str(unsigned char *str, int nr)
+{
+	(MACHINE_IS_VM) ? ASCEBC(str, nr) : ASCEBC_500(str, nr);
+}
+
+static inline struct gds_vector *
+sclp_find_gds_vector(void *start, void *end, u16 id)
+{
+	struct gds_vector *v;
+
+	for (v = start; (void *) v < end; v = (void *) v + v->length)
+		if (v->gds_id == id)
+			return v;
+	return NULL;
+}
+
+static inline struct gds_subvector *
+sclp_find_gds_subvector(void *start, void *end, u8 key)
+{
+	struct gds_subvector *sv;
+
+	for (sv = start; (void *) sv < end; sv = (void *) sv + sv->length)
+		if (sv->key == key)
+			return sv;
+	return NULL;
+}
+
+#endif	 /* __SCLP_H__ */

drivers/s390/char/sclp_async.c

@@ -0,0 +1,211 @@
+/*
+ * Enable Asynchronous Notification via SCLP.
+ *
+ * Copyright IBM Corp. 2009
+ * Author(s): Hans-Joachim Picht <hans@linux.vnet.ibm.com>
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/stat.h>
+#include <linux/string.h>
+#include <linux/slab.h>
+#include <linux/ctype.h>
+#include <linux/kmod.h>
+#include <linux/err.h>
+#include <linux/errno.h>
+#include <linux/proc_fs.h>
+#include <linux/sysctl.h>
+#include <linux/utsname.h>
+#include "sclp.h"
+
+static int callhome_enabled;
+static struct sclp_req *request;
+static struct sclp_async_sccb *sccb;
+static int sclp_async_send_wait(char *message);
+static struct ctl_table_header *callhome_sysctl_header;
+static DEFINE_SPINLOCK(sclp_async_lock);
+#define SCLP_NORMAL_WRITE	0x00
+
+struct async_evbuf {
+	struct evbuf_header header;
+	u64 reserved;
+	u8 rflags;
+	u8 empty;
+	u8 rtype;
+	u8 otype;
+	char comp_id[12];
+	char data[3000]; /* there is still some space left */
+} __attribute__((packed));
+
+struct sclp_async_sccb {
+	struct sccb_header header;
+	struct async_evbuf evbuf;
+} __attribute__((packed));
+
+static struct sclp_register sclp_async_register = {
+	.send_mask = EVTYP_ASYNC_MASK,
+};
+
+static int call_home_on_panic(struct notifier_block *self,
+			      unsigned long event, void *data)
+{
+	strncat(data, init_utsname()->nodename,
+		sizeof(init_utsname()->nodename));
+	sclp_async_send_wait(data);
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block call_home_panic_nb = {
+	.notifier_call = call_home_on_panic,
+	.priority = INT_MAX,
+};
+
+static int proc_handler_callhome(struct ctl_table *ctl, int write,
+				 void __user *buffer, size_t *count,
+				 loff_t *ppos)
+{
+	unsigned long val;
+	int len, rc;
+	char buf[3];
+
+	if (!*count || (*ppos && !write)) {
+		*count = 0;
+		return 0;
+	}
+	if (!write) {
+		len = snprintf(buf, sizeof(buf), "%d\n", callhome_enabled);
+		rc = copy_to_user(buffer, buf, sizeof(buf));
+		if (rc != 0)
+			return -EFAULT;
+	} else {
+		len = *count;
+		rc = kstrtoul_from_user(buffer, len, 0, &val);
+		if (rc)
+			return rc;
+		if (val != 0 && val != 1)
+			return -EINVAL;
+		callhome_enabled = val;
+	}
+	*count = len;
+	*ppos += len;
+	return 0;
+}
+
+static struct ctl_table callhome_table[] = {
+	{
+		.procname	= "callhome",
+		.mode		= 0644,
+		.proc_handler	= proc_handler_callhome,
+	},
+	{}
+};
+
+static struct ctl_table kern_dir_table[] = {
+	{
+		.procname	= "kernel",
+		.maxlen		= 0,
+		.mode		= 0555,
+		.child		= callhome_table,
+	},
+	{}
+};
+
+/*
+ * Function used to transfer asynchronous notification
+ * records which waits for send completion
+ */
+static int sclp_async_send_wait(char *message)
+{
+	struct async_evbuf *evb;
+	int rc;
+	unsigned long flags;
+
+	if (!callhome_enabled)
+		return 0;
+	sccb->evbuf.header.type = EVTYP_ASYNC;
+	sccb->evbuf.rtype = 0xA5;
+	sccb->evbuf.otype = 0x00;
+	evb = &sccb->evbuf;
+	request->command = SCLP_CMDW_WRITE_EVENT_DATA;
+	request->sccb = sccb;
+	request->status = SCLP_REQ_FILLED;
+	strncpy(sccb->evbuf.data, message, sizeof(sccb->evbuf.data));
+	/*
+	 * Retain Queue
+	 * e.g. 5639CC140 500 Red Hat RHEL5 Linux for zSeries (RHEL AS)
+	 */
+	strncpy(sccb->evbuf.comp_id, "000000000", sizeof(sccb->evbuf.comp_id));
+	sccb->evbuf.header.length = sizeof(sccb->evbuf);
+	sccb->header.length = sizeof(sccb->evbuf) + sizeof(sccb->header);
+	sccb->header.function_code = SCLP_NORMAL_WRITE;
+	rc = sclp_add_request(request);
+	if (rc)
+		return rc;
+	spin_lock_irqsave(&sclp_async_lock, flags);
+	while (request->status != SCLP_REQ_DONE &&
+		request->status != SCLP_REQ_FAILED) {
+		 sclp_sync_wait();
+	}
+	spin_unlock_irqrestore(&sclp_async_lock, flags);
+	if (request->status != SCLP_REQ_DONE)
+		return -EIO;
+	rc = ((struct sclp_async_sccb *)
+	       request->sccb)->header.response_code;
+	if (rc != 0x0020)
+		return -EIO;
+	if (evb->header.flags != 0x80)
+		return -EIO;
+	return rc;
+}
+
+static int __init sclp_async_init(void)
+{
+	int rc;
+
+	rc = sclp_register(&sclp_async_register);
+	if (rc)
+		return rc;
+	rc = -EOPNOTSUPP;
+	if (!(sclp_async_register.sclp_receive_mask & EVTYP_ASYNC_MASK))
+		goto out_sclp;
+	rc = -ENOMEM;
+	callhome_sysctl_header = register_sysctl_table(kern_dir_table);
+	if (!callhome_sysctl_header)
+		goto out_sclp;
+	request = kzalloc(sizeof(struct sclp_req), GFP_KERNEL);
+	sccb = (struct sclp_async_sccb *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+	if (!request || !sccb)
+		goto out_mem;
+	rc = atomic_notifier_chain_register(&panic_notifier_list,
+					    &call_home_panic_nb);
+	if (!rc)
+		goto out;
+out_mem:
+	kfree(request);
+	free_page((unsigned long) sccb);
+	unregister_sysctl_table(callhome_sysctl_header);
+out_sclp:
+	sclp_unregister(&sclp_async_register);
+out:
+	return rc;
+}
+module_init(sclp_async_init);
+
+static void __exit sclp_async_exit(void)
+{
+	atomic_notifier_chain_unregister(&panic_notifier_list,
+					 &call_home_panic_nb);
+	unregister_sysctl_table(callhome_sysctl_header);
+	sclp_unregister(&sclp_async_register);
+	free_page((unsigned long) sccb);
+	kfree(request);
+}
+module_exit(sclp_async_exit);
+
+MODULE_AUTHOR("Copyright IBM Corp. 2009");
+MODULE_AUTHOR("Hans-Joachim Picht <hans@linux.vnet.ibm.com>");
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("SCLP Asynchronous Notification Records");

drivers/s390/char/sclp_cmd.c

@@ -0,0 +1,716 @@
+/*
+ * Copyright IBM Corp. 2007,2012
+ *
+ * Author(s): Heiko Carstens <heiko.carstens@de.ibm.com>,
+ *	      Peter Oberparleiter <peter.oberparleiter@de.ibm.com>
+ */
+
+#define KMSG_COMPONENT "sclp_cmd"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/completion.h>
+#include <linux/init.h>
+#include <linux/errno.h>
+#include <linux/err.h>
+#include <linux/export.h>
+#include <linux/slab.h>
+#include <linux/string.h>
+#include <linux/mm.h>
+#include <linux/mmzone.h>
+#include <linux/memory.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <asm/ctl_reg.h>
+#include <asm/chpid.h>
+#include <asm/setup.h>
+#include <asm/page.h>
+#include <asm/sclp.h>
+
+#include "sclp.h"
+
+static void sclp_sync_callback(struct sclp_req *req, void *data)
+{
+	struct completion *completion = data;
+
+	complete(completion);
+}
+
+int sclp_sync_request(sclp_cmdw_t cmd, void *sccb)
+{
+	return sclp_sync_request_timeout(cmd, sccb, 0);
+}
+
+int sclp_sync_request_timeout(sclp_cmdw_t cmd, void *sccb, int timeout)
+{
+	struct completion completion;
+	struct sclp_req *request;
+	int rc;
+
+	request = kzalloc(sizeof(*request), GFP_KERNEL);
+	if (!request)
+		return -ENOMEM;
+	if (timeout)
+		request->queue_timeout = timeout;
+	request->command = cmd;
+	request->sccb = sccb;
+	request->status = SCLP_REQ_FILLED;
+	request->callback = sclp_sync_callback;
+	request->callback_data = &completion;
+	init_completion(&completion);
+
+	/* Perform sclp request. */
+	rc = sclp_add_request(request);
+	if (rc)
+		goto out;
+	wait_for_completion(&completion);
+
+	/* Check response. */
+	if (request->status != SCLP_REQ_DONE) {
+		pr_warning("sync request failed (cmd=0x%08x, "
+			   "status=0x%02x)\n", cmd, request->status);
+		rc = -EIO;
+	}
+out:
+	kfree(request);
+	return rc;
+}
+
+/*
+ * CPU configuration related functions.
+ */
+
+#define SCLP_CMDW_READ_CPU_INFO		0x00010001
+#define SCLP_CMDW_CONFIGURE_CPU		0x00110001
+#define SCLP_CMDW_DECONFIGURE_CPU	0x00100001
+
+struct read_cpu_info_sccb {
+	struct	sccb_header header;
+	u16	nr_configured;
+	u16	offset_configured;
+	u16	nr_standby;
+	u16	offset_standby;
+	u8	reserved[4096 - 16];
+} __attribute__((packed, aligned(PAGE_SIZE)));
+
+static void sclp_fill_cpu_info(struct sclp_cpu_info *info,
+			       struct read_cpu_info_sccb *sccb)
+{
+	char *page = (char *) sccb;
+
+	memset(info, 0, sizeof(*info));
+	info->configured = sccb->nr_configured;
+	info->standby = sccb->nr_standby;
+	info->combined = sccb->nr_configured + sccb->nr_standby;
+	info->has_cpu_type = sclp_fac84 & 0x1;
+	memcpy(&info->cpu, page + sccb->offset_configured,
+	       info->combined * sizeof(struct sclp_cpu_entry));
+}
+
+int sclp_get_cpu_info(struct sclp_cpu_info *info)
+{
+	int rc;
+	struct read_cpu_info_sccb *sccb;
+
+	if (!SCLP_HAS_CPU_INFO)
+		return -EOPNOTSUPP;
+	sccb = (void *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+	if (!sccb)
+		return -ENOMEM;
+	sccb->header.length = sizeof(*sccb);
+	rc = sclp_sync_request_timeout(SCLP_CMDW_READ_CPU_INFO, sccb,
+				       SCLP_QUEUE_INTERVAL);
+	if (rc)
+		goto out;
+	if (sccb->header.response_code != 0x0010) {
+		pr_warning("readcpuinfo failed (response=0x%04x)\n",
+			   sccb->header.response_code);
+		rc = -EIO;
+		goto out;
+	}
+	sclp_fill_cpu_info(info, sccb);
+out:
+	free_page((unsigned long) sccb);
+	return rc;
+}
+
+struct cpu_configure_sccb {
+	struct sccb_header header;
+} __attribute__((packed, aligned(8)));
+
+static int do_cpu_configure(sclp_cmdw_t cmd)
+{
+	struct cpu_configure_sccb *sccb;
+	int rc;
+
+	if (!SCLP_HAS_CPU_RECONFIG)
+		return -EOPNOTSUPP;
+	/*
+	 * This is not going to cross a page boundary since we force
+	 * kmalloc to have a minimum alignment of 8 bytes on s390.
+	 */
+	sccb = kzalloc(sizeof(*sccb), GFP_KERNEL | GFP_DMA);
+	if (!sccb)
+		return -ENOMEM;
+	sccb->header.length = sizeof(*sccb);
+	rc = sclp_sync_request_timeout(cmd, sccb, SCLP_QUEUE_INTERVAL);
+	if (rc)
+		goto out;
+	switch (sccb->header.response_code) {
+	case 0x0020:
+	case 0x0120:
+		break;
+	default:
+		pr_warning("configure cpu failed (cmd=0x%08x, "
+			   "response=0x%04x)\n", cmd,
+			   sccb->header.response_code);
+		rc = -EIO;
+		break;
+	}
+out:
+	kfree(sccb);
+	return rc;
+}
+
+int sclp_cpu_configure(u8 cpu)
+{
+	return do_cpu_configure(SCLP_CMDW_CONFIGURE_CPU | cpu << 8);
+}
+
+int sclp_cpu_deconfigure(u8 cpu)
+{
+	return do_cpu_configure(SCLP_CMDW_DECONFIGURE_CPU | cpu << 8);
+}
+
+#ifdef CONFIG_MEMORY_HOTPLUG
+
+static DEFINE_MUTEX(sclp_mem_mutex);
+static LIST_HEAD(sclp_mem_list);
+static u8 sclp_max_storage_id;
+static unsigned long sclp_storage_ids[256 / BITS_PER_LONG];
+static int sclp_mem_state_changed;
+
+struct memory_increment {
+	struct list_head list;
+	u16 rn;
+	int standby;
+};
+
+struct assign_storage_sccb {
+	struct sccb_header header;
+	u16 rn;
+} __packed;
+
+int arch_get_memory_phys_device(unsigned long start_pfn)
+{
+	if (!sclp_rzm)
+		return 0;
+	return PFN_PHYS(start_pfn) >> ilog2(sclp_rzm);
+}
+
+static unsigned long long rn2addr(u16 rn)
+{
+	return (unsigned long long) (rn - 1) * sclp_rzm;
+}
+
+static int do_assign_storage(sclp_cmdw_t cmd, u16 rn)
+{
+	struct assign_storage_sccb *sccb;
+	int rc;
+
+	sccb = (void *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+	if (!sccb)
+		return -ENOMEM;
+	sccb->header.length = PAGE_SIZE;
+	sccb->rn = rn;
+	rc = sclp_sync_request_timeout(cmd, sccb, SCLP_QUEUE_INTERVAL);
+	if (rc)
+		goto out;
+	switch (sccb->header.response_code) {
+	case 0x0020:
+	case 0x0120:
+		break;
+	default:
+		pr_warning("assign storage failed (cmd=0x%08x, "
+			   "response=0x%04x, rn=0x%04x)\n", cmd,
+			   sccb->header.response_code, rn);
+		rc = -EIO;
+		break;
+	}
+out:
+	free_page((unsigned long) sccb);
+	return rc;
+}
+
+static int sclp_assign_storage(u16 rn)
+{
+	unsigned long long start;
+	int rc;
+
+	rc = do_assign_storage(0x000d0001, rn);
+	if (rc)
+		return rc;
+	start = rn2addr(rn);
+	storage_key_init_range(start, start + sclp_rzm);
+	return 0;
+}
+
+static int sclp_unassign_storage(u16 rn)
+{
+	return do_assign_storage(0x000c0001, rn);
+}
+
+struct attach_storage_sccb {
+	struct sccb_header header;
+	u16 :16;
+	u16 assigned;
+	u32 :32;
+	u32 entries[0];
+} __packed;
+
+static int sclp_attach_storage(u8 id)
+{
+	struct attach_storage_sccb *sccb;
+	int rc;
+	int i;
+
+	sccb = (void *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+	if (!sccb)
+		return -ENOMEM;
+	sccb->header.length = PAGE_SIZE;
+	rc = sclp_sync_request_timeout(0x00080001 | id << 8, sccb,
+				       SCLP_QUEUE_INTERVAL);
+	if (rc)
+		goto out;
+	switch (sccb->header.response_code) {
+	case 0x0020:
+		set_bit(id, sclp_storage_ids);
+		for (i = 0; i < sccb->assigned; i++) {
+			if (sccb->entries[i])
+				sclp_unassign_storage(sccb->entries[i] >> 16);
+		}
+		break;
+	default:
+		rc = -EIO;
+		break;
+	}
+out:
+	free_page((unsigned long) sccb);
+	return rc;
+}
+
+static int sclp_mem_change_state(unsigned long start, unsigned long size,
+				 int online)
+{
+	struct memory_increment *incr;
+	unsigned long long istart;
+	int rc = 0;
+
+	list_for_each_entry(incr, &sclp_mem_list, list) {
+		istart = rn2addr(incr->rn);
+		if (start + size - 1 < istart)
+			break;
+		if (start > istart + sclp_rzm - 1)
+			continue;
+		if (online)
+			rc |= sclp_assign_storage(incr->rn);
+		else
+			sclp_unassign_storage(incr->rn);
+	}
+	return rc ? -EIO : 0;
+}
+
+static int sclp_mem_notifier(struct notifier_block *nb,
+			     unsigned long action, void *data)
+{
+	unsigned long start, size;
+	struct memory_notify *arg;
+	unsigned char id;
+	int rc = 0;
+
+	arg = data;
+	start = arg->start_pfn << PAGE_SHIFT;
+	size = arg->nr_pages << PAGE_SHIFT;
+	mutex_lock(&sclp_mem_mutex);
+	for_each_clear_bit(id, sclp_storage_ids, sclp_max_storage_id + 1)
+		sclp_attach_storage(id);
+	switch (action) {
+	case MEM_ONLINE:
+	case MEM_GOING_OFFLINE:
+	case MEM_CANCEL_OFFLINE:
+		break;
+	case MEM_GOING_ONLINE:
+		rc = sclp_mem_change_state(start, size, 1);
+		break;
+	case MEM_CANCEL_ONLINE:
+		sclp_mem_change_state(start, size, 0);
+		break;
+	case MEM_OFFLINE:
+		sclp_mem_change_state(start, size, 0);
+		break;
+	default:
+		rc = -EINVAL;
+		break;
+	}
+	if (!rc)
+		sclp_mem_state_changed = 1;
+	mutex_unlock(&sclp_mem_mutex);
+	return rc ? NOTIFY_BAD : NOTIFY_OK;
+}
+
+static struct notifier_block sclp_mem_nb = {
+	.notifier_call = sclp_mem_notifier,
+};
+
+static void __init add_memory_merged(u16 rn)
+{
+	static u16 first_rn, num;
+	unsigned long long start, size;
+
+	if (rn && first_rn && (first_rn + num == rn)) {
+		num++;
+		return;
+	}
+	if (!first_rn)
+		goto skip_add;
+	start = rn2addr(first_rn);
+	size = (unsigned long long) num * sclp_rzm;
+	if (start >= VMEM_MAX_PHYS)
+		goto skip_add;
+	if (start + size > VMEM_MAX_PHYS)
+		size = VMEM_MAX_PHYS - start;
+	if (memory_end_set && (start >= memory_end))
+		goto skip_add;
+	if (memory_end_set && (start + size > memory_end))
+		size = memory_end - start;
+	add_memory(0, start, size);
+skip_add:
+	first_rn = rn;
+	num = 1;
+}
+
+static void __init sclp_add_standby_memory(void)
+{
+	struct memory_increment *incr;
+
+	list_for_each_entry(incr, &sclp_mem_list, list)
+		if (incr->standby)
+			add_memory_merged(incr->rn);
+	add_memory_merged(0);
+}
+
+static void __init insert_increment(u16 rn, int standby, int assigned)
+{
+	struct memory_increment *incr, *new_incr;
+	struct list_head *prev;
+	u16 last_rn;
+
+	new_incr = kzalloc(sizeof(*new_incr), GFP_KERNEL);
+	if (!new_incr)
+		return;
+	new_incr->rn = rn;
+	new_incr->standby = standby;
+	last_rn = 0;
+	prev = &sclp_mem_list;
+	list_for_each_entry(incr, &sclp_mem_list, list) {
+		if (assigned && incr->rn > rn)
+			break;
+		if (!assigned && incr->rn - last_rn > 1)
+			break;
+		last_rn = incr->rn;
+		prev = &incr->list;
+	}
+	if (!assigned)
+		new_incr->rn = last_rn + 1;
+	if (new_incr->rn > sclp_rnmax) {
+		kfree(new_incr);
+		return;
+	}
+	list_add(&new_incr->list, prev);
+}
+
+static int sclp_mem_freeze(struct device *dev)
+{
+	if (!sclp_mem_state_changed)
+		return 0;
+	pr_err("Memory hotplug state changed, suspend refused.\n");
+	return -EPERM;
+}
+
+struct read_storage_sccb {
+	struct sccb_header header;
+	u16 max_id;
+	u16 assigned;
+	u16 standby;
+	u16 :16;
+	u32 entries[0];
+} __packed;
+
+static const struct dev_pm_ops sclp_mem_pm_ops = {
+	.freeze		= sclp_mem_freeze,
+};
+
+static struct platform_driver sclp_mem_pdrv = {
+	.driver = {
+		.name	= "sclp_mem",
+		.pm	= &sclp_mem_pm_ops,
+	},
+};
+
+static int __init sclp_detect_standby_memory(void)
+{
+	struct platform_device *sclp_pdev;
+	struct read_storage_sccb *sccb;
+	int i, id, assigned, rc;
+
+	if (OLDMEM_BASE) /* No standby memory in kdump mode */
+		return 0;
+	if ((sclp_facilities & 0xe00000000000ULL) != 0xe00000000000ULL)
+		return 0;
+	rc = -ENOMEM;
+	sccb = (void *) __get_free_page(GFP_KERNEL | GFP_DMA);
+	if (!sccb)
+		goto out;
+	assigned = 0;
+	for (id = 0; id <= sclp_max_storage_id; id++) {
+		memset(sccb, 0, PAGE_SIZE);
+		sccb->header.length = PAGE_SIZE;
+		rc = sclp_sync_request(0x00040001 | id << 8, sccb);
+		if (rc)
+			goto out;
+		switch (sccb->header.response_code) {
+		case 0x0010:
+			set_bit(id, sclp_storage_ids);
+			for (i = 0; i < sccb->assigned; i++) {
+				if (!sccb->entries[i])
+					continue;
+				assigned++;
+				insert_increment(sccb->entries[i] >> 16, 0, 1);
+			}
+			break;
+		case 0x0310:
+			break;
+		case 0x0410:
+			for (i = 0; i < sccb->assigned; i++) {
+				if (!sccb->entries[i])
+					continue;
+				assigned++;
+				insert_increment(sccb->entries[i] >> 16, 1, 1);
+			}
+			break;
+		default:
+			rc = -EIO;
+			break;
+		}
+		if (!rc)
+			sclp_max_storage_id = sccb->max_id;
+	}
+	if (rc || list_empty(&sclp_mem_list))
+		goto out;
+	for (i = 1; i <= sclp_rnmax - assigned; i++)
+		insert_increment(0, 1, 0);
+	rc = register_memory_notifier(&sclp_mem_nb);
+	if (rc)
+		goto out;
+	rc = platform_driver_register(&sclp_mem_pdrv);
+	if (rc)
+		goto out;
+	sclp_pdev = platform_device_register_simple("sclp_mem", -1, NULL, 0);
+	rc = PTR_ERR_OR_ZERO(sclp_pdev);
+	if (rc)
+		goto out_driver;
+	sclp_add_standby_memory();
+	goto out;
+out_driver:
+	platform_driver_unregister(&sclp_mem_pdrv);
+out:
+	free_page((unsigned long) sccb);
+	return rc;
+}
+__initcall(sclp_detect_standby_memory);
+
+#endif /* CONFIG_MEMORY_HOTPLUG */
+
+/*
+ * PCI I/O adapter configuration related functions.
+ */
+#define SCLP_CMDW_CONFIGURE_PCI			0x001a0001
+#define SCLP_CMDW_DECONFIGURE_PCI		0x001b0001
+
+#define SCLP_RECONFIG_PCI_ATPYE			2
+
+struct pci_cfg_sccb {
+	struct sccb_header header;
+	u8 atype;		/* adapter type */
+	u8 reserved1;
+	u16 reserved2;
+	u32 aid;		/* adapter identifier */
+} __packed;
+
+static int do_pci_configure(sclp_cmdw_t cmd, u32 fid)
+{
+	struct pci_cfg_sccb *sccb;
+	int rc;
+
+	if (!SCLP_HAS_PCI_RECONFIG)
+		return -EOPNOTSUPP;
+
+	sccb = (struct pci_cfg_sccb *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+	if (!sccb)
+		return -ENOMEM;
+
+	sccb->header.length = PAGE_SIZE;
+	sccb->atype = SCLP_RECONFIG_PCI_ATPYE;
+	sccb->aid = fid;
+	rc = sclp_sync_request(cmd, sccb);
+	if (rc)
+		goto out;
+	switch (sccb->header.response_code) {
+	case 0x0020:
+	case 0x0120:
+		break;
+	default:
+		pr_warn("configure PCI I/O adapter failed: cmd=0x%08x  response=0x%04x\n",
+			cmd, sccb->header.response_code);
+		rc = -EIO;
+		break;
+	}
+out:
+	free_page((unsigned long) sccb);
+	return rc;
+}
+
+int sclp_pci_configure(u32 fid)
+{
+	return do_pci_configure(SCLP_CMDW_CONFIGURE_PCI, fid);
+}
+EXPORT_SYMBOL(sclp_pci_configure);
+
+int sclp_pci_deconfigure(u32 fid)
+{
+	return do_pci_configure(SCLP_CMDW_DECONFIGURE_PCI, fid);
+}
+EXPORT_SYMBOL(sclp_pci_deconfigure);
+
+/*
+ * Channel path configuration related functions.
+ */
+
+#define SCLP_CMDW_CONFIGURE_CHPATH		0x000f0001
+#define SCLP_CMDW_DECONFIGURE_CHPATH		0x000e0001
+#define SCLP_CMDW_READ_CHPATH_INFORMATION	0x00030001
+
+struct chp_cfg_sccb {
+	struct sccb_header header;
+	u8 ccm;
+	u8 reserved[6];
+	u8 cssid;
+} __attribute__((packed));
+
+static int do_chp_configure(sclp_cmdw_t cmd)
+{
+	struct chp_cfg_sccb *sccb;
+	int rc;
+
+	if (!SCLP_HAS_CHP_RECONFIG)
+		return -EOPNOTSUPP;
+	/* Prepare sccb. */
+	sccb = (struct chp_cfg_sccb *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+	if (!sccb)
+		return -ENOMEM;
+	sccb->header.length = sizeof(*sccb);
+	rc = sclp_sync_request(cmd, sccb);
+	if (rc)
+		goto out;
+	switch (sccb->header.response_code) {
+	case 0x0020:
+	case 0x0120:
+	case 0x0440:
+	case 0x0450:
+		break;
+	default:
+		pr_warning("configure channel-path failed "
+			   "(cmd=0x%08x, response=0x%04x)\n", cmd,
+			   sccb->header.response_code);
+		rc = -EIO;
+		break;
+	}
+out:
+	free_page((unsigned long) sccb);
+	return rc;
+}
+
+/**
+ * sclp_chp_configure - perform configure channel-path sclp command
+ * @chpid: channel-path ID
+ *
+ * Perform configure channel-path command sclp command for specified chpid.
+ * Return 0 after command successfully finished, non-zero otherwise.
+ */
+int sclp_chp_configure(struct chp_id chpid)
+{
+	return do_chp_configure(SCLP_CMDW_CONFIGURE_CHPATH | chpid.id << 8);
+}
+
+/**
+ * sclp_chp_deconfigure - perform deconfigure channel-path sclp command
+ * @chpid: channel-path ID
+ *
+ * Perform deconfigure channel-path command sclp command for specified chpid
+ * and wait for completion. On success return 0. Return non-zero otherwise.
+ */
+int sclp_chp_deconfigure(struct chp_id chpid)
+{
+	return do_chp_configure(SCLP_CMDW_DECONFIGURE_CHPATH | chpid.id << 8);
+}
+
+struct chp_info_sccb {
+	struct sccb_header header;
+	u8 recognized[SCLP_CHP_INFO_MASK_SIZE];
+	u8 standby[SCLP_CHP_INFO_MASK_SIZE];
+	u8 configured[SCLP_CHP_INFO_MASK_SIZE];
+	u8 ccm;
+	u8 reserved[6];
+	u8 cssid;
+} __attribute__((packed));
+
+/**
+ * sclp_chp_read_info - perform read channel-path information sclp command
+ * @info: resulting channel-path information data
+ *
+ * Perform read channel-path information sclp command and wait for completion.
+ * On success, store channel-path information in @info and return 0. Return
+ * non-zero otherwise.
+ */
+int sclp_chp_read_info(struct sclp_chp_info *info)
+{
+	struct chp_info_sccb *sccb;
+	int rc;
+
+	if (!SCLP_HAS_CHP_INFO)
+		return -EOPNOTSUPP;
+	/* Prepare sccb. */
+	sccb = (struct chp_info_sccb *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+	if (!sccb)
+		return -ENOMEM;
+	sccb->header.length = sizeof(*sccb);
+	rc = sclp_sync_request(SCLP_CMDW_READ_CHPATH_INFORMATION, sccb);
+	if (rc)
+		goto out;
+	if (sccb->header.response_code != 0x0010) {
+		pr_warning("read channel-path info failed "
+			   "(response=0x%04x)\n", sccb->header.response_code);
+		rc = -EIO;
+		goto out;
+	}
+	memcpy(info->recognized, sccb->recognized, SCLP_CHP_INFO_MASK_SIZE);
+	memcpy(info->standby, sccb->standby, SCLP_CHP_INFO_MASK_SIZE);
+	memcpy(info->configured, sccb->configured, SCLP_CHP_INFO_MASK_SIZE);
+out:
+	free_page((unsigned long) sccb);
+	return rc;
+}
+
+bool sclp_has_sprp(void)
+{
+	return !!(sclp_fac84 & 0x2);
+}

drivers/s390/char/sclp_con.c

@@ -0,0 +1,356 @@
+/*
+ * SCLP line mode console driver
+ *
+ * Copyright IBM Corp. 1999, 2009
+ * Author(s): Martin Peschke <mpeschke@de.ibm.com>
+ *	      Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#include <linux/kmod.h>
+#include <linux/console.h>
+#include <linux/init.h>
+#include <linux/timer.h>
+#include <linux/jiffies.h>
+#include <linux/termios.h>
+#include <linux/err.h>
+#include <linux/reboot.h>
+#include <linux/gfp.h>
+
+#include "sclp.h"
+#include "sclp_rw.h"
+#include "sclp_tty.h"
+
+#define sclp_console_major 4		/* TTYAUX_MAJOR */
+#define sclp_console_minor 64
+#define sclp_console_name  "ttyS"
+
+/* Lock to guard over changes to global variables */
+static spinlock_t sclp_con_lock;
+/* List of free pages that can be used for console output buffering */
+static struct list_head sclp_con_pages;
+/* List of full struct sclp_buffer structures ready for output */
+static struct list_head sclp_con_outqueue;
+/* Pointer to current console buffer */
+static struct sclp_buffer *sclp_conbuf;
+/* Timer for delayed output of console messages */
+static struct timer_list sclp_con_timer;
+/* Suspend mode flag */
+static int sclp_con_suspended;
+/* Flag that output queue is currently running */
+static int sclp_con_queue_running;
+
+/* Output format for console messages */
+static unsigned short sclp_con_columns;
+static unsigned short sclp_con_width_htab;
+
+static void
+sclp_conbuf_callback(struct sclp_buffer *buffer, int rc)
+{
+	unsigned long flags;
+	void *page;
+
+	do {
+		page = sclp_unmake_buffer(buffer);
+		spin_lock_irqsave(&sclp_con_lock, flags);
+
+		/* Remove buffer from outqueue */
+		list_del(&buffer->list);
+		list_add_tail((struct list_head *) page, &sclp_con_pages);
+
+		/* Check if there is a pending buffer on the out queue. */
+		buffer = NULL;
+		if (!list_empty(&sclp_con_outqueue))
+			buffer = list_first_entry(&sclp_con_outqueue,
+						  struct sclp_buffer, list);
+		if (!buffer || sclp_con_suspended) {
+			sclp_con_queue_running = 0;
+			spin_unlock_irqrestore(&sclp_con_lock, flags);
+			break;
+		}
+		spin_unlock_irqrestore(&sclp_con_lock, flags);
+	} while (sclp_emit_buffer(buffer, sclp_conbuf_callback));
+}
+
+/*
+ * Finalize and emit first pending buffer.
+ */
+static void sclp_conbuf_emit(void)
+{
+	struct sclp_buffer* buffer;
+	unsigned long flags;
+	int rc;
+
+	spin_lock_irqsave(&sclp_con_lock, flags);
+	if (sclp_conbuf)
+		list_add_tail(&sclp_conbuf->list, &sclp_con_outqueue);
+	sclp_conbuf = NULL;
+	if (sclp_con_queue_running || sclp_con_suspended)
+		goto out_unlock;
+	if (list_empty(&sclp_con_outqueue))
+		goto out_unlock;
+	buffer = list_first_entry(&sclp_con_outqueue, struct sclp_buffer,
+				  list);
+	sclp_con_queue_running = 1;
+	spin_unlock_irqrestore(&sclp_con_lock, flags);
+
+	rc = sclp_emit_buffer(buffer, sclp_conbuf_callback);
+	if (rc)
+		sclp_conbuf_callback(buffer, rc);
+	return;
+out_unlock:
+	spin_unlock_irqrestore(&sclp_con_lock, flags);
+}
+
+/*
+ * Wait until out queue is empty
+ */
+static void sclp_console_sync_queue(void)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&sclp_con_lock, flags);
+	if (timer_pending(&sclp_con_timer))
+		del_timer(&sclp_con_timer);
+	while (sclp_con_queue_running) {
+		spin_unlock_irqrestore(&sclp_con_lock, flags);
+		sclp_sync_wait();
+		spin_lock_irqsave(&sclp_con_lock, flags);
+	}
+	spin_unlock_irqrestore(&sclp_con_lock, flags);
+}
+
+/*
+ * When this routine is called from the timer then we flush the
+ * temporary write buffer without further waiting on a final new line.
+ */
+static void
+sclp_console_timeout(unsigned long data)
+{
+	sclp_conbuf_emit();
+}
+
+/*
+ * Drop oldest console buffer if sclp_con_drop is set
+ */
+static int
+sclp_console_drop_buffer(void)
+{
+	struct list_head *list;
+	struct sclp_buffer *buffer;
+	void *page;
+
+	if (!sclp_console_drop)
+		return 0;
+	list = sclp_con_outqueue.next;
+	if (sclp_con_queue_running)
+		/* The first element is in I/O */
+		list = list->next;
+	if (list == &sclp_con_outqueue)
+		return 0;
+	list_del(list);
+	buffer = list_entry(list, struct sclp_buffer, list);
+	page = sclp_unmake_buffer(buffer);
+	list_add_tail((struct list_head *) page, &sclp_con_pages);
+	return 1;
+}
+
+/*
+ * Writes the given message to S390 system console
+ */
+static void
+sclp_console_write(struct console *console, const char *message,
+		   unsigned int count)
+{
+	unsigned long flags;
+	void *page;
+	int written;
+
+	if (count == 0)
+		return;
+	spin_lock_irqsave(&sclp_con_lock, flags);
+	/*
+	 * process escape characters, write message into buffer,
+	 * send buffer to SCLP
+	 */
+	do {
+		/* make sure we have a console output buffer */
+		if (sclp_conbuf == NULL) {
+			if (list_empty(&sclp_con_pages))
+				sclp_console_full++;
+			while (list_empty(&sclp_con_pages)) {
+				if (sclp_con_suspended)
+					goto out;
+				if (sclp_console_drop_buffer())
+					break;
+				spin_unlock_irqrestore(&sclp_con_lock, flags);
+				sclp_sync_wait();
+				spin_lock_irqsave(&sclp_con_lock, flags);
+			}
+			page = sclp_con_pages.next;
+			list_del((struct list_head *) page);
+			sclp_conbuf = sclp_make_buffer(page, sclp_con_columns,
+						       sclp_con_width_htab);
+		}
+		/* try to write the string to the current output buffer */
+		written = sclp_write(sclp_conbuf, (const unsigned char *)
+				     message, count);
+		if (written == count)
+			break;
+		/*
+		 * Not all characters could be written to the current
+		 * output buffer. Emit the buffer, create a new buffer
+		 * and then output the rest of the string.
+		 */
+		spin_unlock_irqrestore(&sclp_con_lock, flags);
+		sclp_conbuf_emit();
+		spin_lock_irqsave(&sclp_con_lock, flags);
+		message += written;
+		count -= written;
+	} while (count > 0);
+	/* Setup timer to output current console buffer after 1/10 second */
+	if (sclp_conbuf != NULL && sclp_chars_in_buffer(sclp_conbuf) != 0 &&
+	    !timer_pending(&sclp_con_timer)) {
+		init_timer(&sclp_con_timer);
+		sclp_con_timer.function = sclp_console_timeout;
+		sclp_con_timer.data = 0UL;
+		sclp_con_timer.expires = jiffies + HZ/10;
+		add_timer(&sclp_con_timer);
+	}
+out:
+	spin_unlock_irqrestore(&sclp_con_lock, flags);
+}
+
+static struct tty_driver *
+sclp_console_device(struct console *c, int *index)
+{
+	*index = c->index;
+	return sclp_tty_driver;
+}
+
+/*
+ * Make sure that all buffers will be flushed to the SCLP.
+ */
+static void
+sclp_console_flush(void)
+{
+	sclp_conbuf_emit();
+	sclp_console_sync_queue();
+}
+
+/*
+ * Resume console: If there are cached messages, emit them.
+ */
+static void sclp_console_resume(void)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&sclp_con_lock, flags);
+	sclp_con_suspended = 0;
+	spin_unlock_irqrestore(&sclp_con_lock, flags);
+	sclp_conbuf_emit();
+}
+
+/*
+ * Suspend console: Set suspend flag and flush console
+ */
+static void sclp_console_suspend(void)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&sclp_con_lock, flags);
+	sclp_con_suspended = 1;
+	spin_unlock_irqrestore(&sclp_con_lock, flags);
+	sclp_console_flush();
+}
+
+static int sclp_console_notify(struct notifier_block *self,
+			       unsigned long event, void *data)
+{
+	sclp_console_flush();
+	return NOTIFY_OK;
+}
+
+static struct notifier_block on_panic_nb = {
+	.notifier_call = sclp_console_notify,
+	.priority = SCLP_PANIC_PRIO_CLIENT,
+};
+
+static struct notifier_block on_reboot_nb = {
+	.notifier_call = sclp_console_notify,
+	.priority = 1,
+};
+
+/*
+ * used to register the SCLP console to the kernel and to
+ * give printk necessary information
+ */
+static struct console sclp_console =
+{
+	.name = sclp_console_name,
+	.write = sclp_console_write,
+	.device = sclp_console_device,
+	.flags = CON_PRINTBUFFER,
+	.index = 0 /* ttyS0 */
+};
+
+/*
+ * This function is called for SCLP suspend and resume events.
+ */
+void sclp_console_pm_event(enum sclp_pm_event sclp_pm_event)
+{
+	switch (sclp_pm_event) {
+	case SCLP_PM_EVENT_FREEZE:
+		sclp_console_suspend();
+		break;
+	case SCLP_PM_EVENT_RESTORE:
+	case SCLP_PM_EVENT_THAW:
+		sclp_console_resume();
+		break;
+	}
+}
+
+/*
+ * called by console_init() in drivers/char/tty_io.c at boot-time.
+ */
+static int __init
+sclp_console_init(void)
+{
+	void *page;
+	int i;
+	int rc;
+
+	if (!CONSOLE_IS_SCLP)
+		return 0;
+	rc = sclp_rw_init();
+	if (rc)
+		return rc;
+	/* Allocate pages for output buffering */
+	INIT_LIST_HEAD(&sclp_con_pages);
+	for (i = 0; i < sclp_console_pages; i++) {
+		page = (void *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+		list_add_tail(page, &sclp_con_pages);
+	}
+	INIT_LIST_HEAD(&sclp_con_outqueue);
+	spin_lock_init(&sclp_con_lock);
+	sclp_conbuf = NULL;
+	init_timer(&sclp_con_timer);
+
+	/* Set output format */
+	if (MACHINE_IS_VM)
+		/*
+		 * save 4 characters for the CPU number
+		 * written at start of each line by VM/CP
+		 */
+		sclp_con_columns = 76;
+	else
+		sclp_con_columns = 80;
+	sclp_con_width_htab = 8;
+
+	/* enable printk-access to this driver */
+	atomic_notifier_chain_register(&panic_notifier_list, &on_panic_nb);
+	register_reboot_notifier(&on_reboot_nb);
+	register_console(&sclp_console);
+	return 0;
+}
+
+console_initcall(sclp_console_init);

drivers/s390/char/sclp_config.c

@@ -0,0 +1,77 @@
+/*
+ *    Copyright IBM Corp. 2007
+ *    Author(s): Heiko Carstens <heiko.carstens@de.ibm.com>
+ */
+
+#define KMSG_COMPONENT "sclp_config"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/init.h>
+#include <linux/errno.h>
+#include <linux/cpu.h>
+#include <linux/device.h>
+#include <linux/workqueue.h>
+#include <asm/smp.h>
+
+#include "sclp.h"
+
+struct conf_mgm_data {
+	u8 reserved;
+	u8 ev_qualifier;
+} __attribute__((packed));
+
+#define EV_QUAL_CPU_CHANGE	1
+#define EV_QUAL_CAP_CHANGE	3
+
+static struct work_struct sclp_cpu_capability_work;
+static struct work_struct sclp_cpu_change_work;
+
+static void sclp_cpu_capability_notify(struct work_struct *work)
+{
+	int cpu;
+	struct device *dev;
+
+	s390_adjust_jiffies();
+	pr_info("CPU capability may have changed\n");
+	get_online_cpus();
+	for_each_online_cpu(cpu) {
+		dev = get_cpu_device(cpu);
+		kobject_uevent(&dev->kobj, KOBJ_CHANGE);
+	}
+	put_online_cpus();
+}
+
+static void __ref sclp_cpu_change_notify(struct work_struct *work)
+{
+	smp_rescan_cpus();
+}
+
+static void sclp_conf_receiver_fn(struct evbuf_header *evbuf)
+{
+	struct conf_mgm_data *cdata;
+
+	cdata = (struct conf_mgm_data *)(evbuf + 1);
+	switch (cdata->ev_qualifier) {
+	case EV_QUAL_CPU_CHANGE:
+		schedule_work(&sclp_cpu_change_work);
+		break;
+	case EV_QUAL_CAP_CHANGE:
+		schedule_work(&sclp_cpu_capability_work);
+		break;
+	}
+}
+
+static struct sclp_register sclp_conf_register =
+{
+	.receive_mask = EVTYP_CONFMGMDATA_MASK,
+	.receiver_fn  = sclp_conf_receiver_fn,
+};
+
+static int __init sclp_conf_init(void)
+{
+	INIT_WORK(&sclp_cpu_capability_work, sclp_cpu_capability_notify);
+	INIT_WORK(&sclp_cpu_change_work, sclp_cpu_change_notify);
+	return sclp_register(&sclp_conf_register);
+}
+
+__initcall(sclp_conf_init);

drivers/s390/char/sclp_cpi.c

@@ -0,0 +1,40 @@
+/*
+ *    SCLP control programm identification
+ *
+ *    Copyright IBM Corp. 2001, 2007
+ *    Author(s): Martin Peschke <mpeschke@de.ibm.com>
+ *		 Michael Ernst <mernst@de.ibm.com>
+ */
+
+#include <linux/kmod.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/version.h>
+#include "sclp_cpi_sys.h"
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Identify this operating system instance "
+		   "to the System z hardware");
+MODULE_AUTHOR("Martin Peschke <mpeschke@de.ibm.com>, "
+	      "Michael Ernst <mernst@de.ibm.com>");
+
+static char *system_name = "";
+static char *sysplex_name = "";
+
+module_param(system_name, charp, 0);
+MODULE_PARM_DESC(system_name, "e.g. hostname - max. 8 characters");
+module_param(sysplex_name, charp, 0);
+MODULE_PARM_DESC(sysplex_name, "if applicable - max. 8 characters");
+
+static int __init cpi_module_init(void)
+{
+	return sclp_cpi_set_data(system_name, sysplex_name, "LINUX",
+				 LINUX_VERSION_CODE);
+}
+
+static void __exit cpi_module_exit(void)
+{
+}
+
+module_init(cpi_module_init);
+module_exit(cpi_module_exit);

drivers/s390/char/sclp_cpi_sys.c

@@ -0,0 +1,429 @@
+/*
+ *    SCLP control program identification sysfs interface
+ *
+ *    Copyright IBM Corp. 2001, 2007
+ *    Author(s): Martin Peschke <mpeschke@de.ibm.com>
+ *		 Michael Ernst <mernst@de.ibm.com>
+ */
+
+#define KMSG_COMPONENT "sclp_cpi"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/stat.h>
+#include <linux/device.h>
+#include <linux/string.h>
+#include <linux/ctype.h>
+#include <linux/kmod.h>
+#include <linux/timer.h>
+#include <linux/err.h>
+#include <linux/slab.h>
+#include <linux/completion.h>
+#include <linux/export.h>
+#include <asm/ebcdic.h>
+#include <asm/sclp.h>
+
+#include "sclp.h"
+#include "sclp_rw.h"
+#include "sclp_cpi_sys.h"
+
+#define CPI_LENGTH_NAME 8
+#define CPI_LENGTH_LEVEL 16
+
+static DEFINE_MUTEX(sclp_cpi_mutex);
+
+struct cpi_evbuf {
+	struct evbuf_header header;
+	u8	id_format;
+	u8	reserved0;
+	u8	system_type[CPI_LENGTH_NAME];
+	u64	reserved1;
+	u8	system_name[CPI_LENGTH_NAME];
+	u64	reserved2;
+	u64	system_level;
+	u64	reserved3;
+	u8	sysplex_name[CPI_LENGTH_NAME];
+	u8	reserved4[16];
+} __attribute__((packed));
+
+struct cpi_sccb {
+	struct sccb_header header;
+	struct cpi_evbuf cpi_evbuf;
+} __attribute__((packed));
+
+static struct sclp_register sclp_cpi_event = {
+	.send_mask = EVTYP_CTLPROGIDENT_MASK,
+};
+
+static char system_name[CPI_LENGTH_NAME + 1];
+static char sysplex_name[CPI_LENGTH_NAME + 1];
+static char system_type[CPI_LENGTH_NAME + 1];
+static u64 system_level;
+
+static void set_data(char *field, char *data)
+{
+	memset(field, ' ', CPI_LENGTH_NAME);
+	memcpy(field, data, strlen(data));
+	sclp_ascebc_str(field, CPI_LENGTH_NAME);
+}
+
+static void cpi_callback(struct sclp_req *req, void *data)
+{
+	struct completion *completion = data;
+
+	complete(completion);
+}
+
+static struct sclp_req *cpi_prepare_req(void)
+{
+	struct sclp_req *req;
+	struct cpi_sccb *sccb;
+	struct cpi_evbuf *evb;
+
+	req = kzalloc(sizeof(struct sclp_req), GFP_KERNEL);
+	if (!req)
+		return ERR_PTR(-ENOMEM);
+	sccb = (struct cpi_sccb *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+	if (!sccb) {
+		kfree(req);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	/* setup SCCB for Control-Program Identification */
+	sccb->header.length = sizeof(struct cpi_sccb);
+	sccb->cpi_evbuf.header.length = sizeof(struct cpi_evbuf);
+	sccb->cpi_evbuf.header.type = 0x0b;
+	evb = &sccb->cpi_evbuf;
+
+	/* set system type */
+	set_data(evb->system_type, system_type);
+
+	/* set system name */
+	set_data(evb->system_name, system_name);
+
+	/* set system level */
+	evb->system_level = system_level;
+
+	/* set sysplex name */
+	set_data(evb->sysplex_name, sysplex_name);
+
+	/* prepare request data structure presented to SCLP driver */
+	req->command = SCLP_CMDW_WRITE_EVENT_DATA;
+	req->sccb = sccb;
+	req->status = SCLP_REQ_FILLED;
+	req->callback = cpi_callback;
+	return req;
+}
+
+static void cpi_free_req(struct sclp_req *req)
+{
+	free_page((unsigned long) req->sccb);
+	kfree(req);
+}
+
+static int cpi_req(void)
+{
+	struct completion completion;
+	struct sclp_req *req;
+	int rc;
+	int response;
+
+	rc = sclp_register(&sclp_cpi_event);
+	if (rc)
+		goto out;
+	if (!(sclp_cpi_event.sclp_receive_mask & EVTYP_CTLPROGIDENT_MASK)) {
+		rc = -EOPNOTSUPP;
+		goto out_unregister;
+	}
+
+	req = cpi_prepare_req();
+	if (IS_ERR(req)) {
+		rc = PTR_ERR(req);
+		goto out_unregister;
+	}
+
+	init_completion(&completion);
+	req->callback_data = &completion;
+
+	/* Add request to sclp queue */
+	rc = sclp_add_request(req);
+	if (rc)
+		goto out_free_req;
+
+	wait_for_completion(&completion);
+
+	if (req->status != SCLP_REQ_DONE) {
+		pr_warning("request failed (status=0x%02x)\n",
+			   req->status);
+		rc = -EIO;
+		goto out_free_req;
+	}
+
+	response = ((struct cpi_sccb *) req->sccb)->header.response_code;
+	if (response != 0x0020) {
+		pr_warning("request failed with response code 0x%x\n",
+			   response);
+		rc = -EIO;
+	}
+
+out_free_req:
+	cpi_free_req(req);
+
+out_unregister:
+	sclp_unregister(&sclp_cpi_event);
+
+out:
+	return rc;
+}
+
+static int check_string(const char *attr, const char *str)
+{
+	size_t len;
+	size_t i;
+
+	len = strlen(str);
+
+	if ((len > 0) && (str[len - 1] == '\n'))
+		len--;
+
+	if (len > CPI_LENGTH_NAME)
+		return -EINVAL;
+
+	for (i = 0; i < len ; i++) {
+		if (isalpha(str[i]) || isdigit(str[i]) ||
+		    strchr("$@# ", str[i]))
+			continue;
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static void set_string(char *attr, const char *value)
+{
+	size_t len;
+	size_t i;
+
+	len = strlen(value);
+
+	if ((len > 0) && (value[len - 1] == '\n'))
+		len--;
+
+	for (i = 0; i < CPI_LENGTH_NAME; i++) {
+		if (i < len)
+			attr[i] = toupper(value[i]);
+		else
+			attr[i] = ' ';
+	}
+}
+
+static ssize_t system_name_show(struct kobject *kobj,
+				struct kobj_attribute *attr, char *page)
+{
+	int rc;
+
+	mutex_lock(&sclp_cpi_mutex);
+	rc = snprintf(page, PAGE_SIZE, "%s\n", system_name);
+	mutex_unlock(&sclp_cpi_mutex);
+	return rc;
+}
+
+static ssize_t system_name_store(struct kobject *kobj,
+				 struct kobj_attribute *attr,
+				 const char *buf,
+	size_t len)
+{
+	int rc;
+
+	rc = check_string("system_name", buf);
+	if (rc)
+		return rc;
+
+	mutex_lock(&sclp_cpi_mutex);
+	set_string(system_name, buf);
+	mutex_unlock(&sclp_cpi_mutex);
+
+	return len;
+}
+
+static struct kobj_attribute system_name_attr =
+	__ATTR(system_name, 0644, system_name_show, system_name_store);
+
+static ssize_t sysplex_name_show(struct kobject *kobj,
+				 struct kobj_attribute *attr, char *page)
+{
+	int rc;
+
+	mutex_lock(&sclp_cpi_mutex);
+	rc = snprintf(page, PAGE_SIZE, "%s\n", sysplex_name);
+	mutex_unlock(&sclp_cpi_mutex);
+	return rc;
+}
+
+static ssize_t sysplex_name_store(struct kobject *kobj,
+				  struct kobj_attribute *attr,
+				  const char *buf,
+	size_t len)
+{
+	int rc;
+
+	rc = check_string("sysplex_name", buf);
+	if (rc)
+		return rc;
+
+	mutex_lock(&sclp_cpi_mutex);
+	set_string(sysplex_name, buf);
+	mutex_unlock(&sclp_cpi_mutex);
+
+	return len;
+}
+
+static struct kobj_attribute sysplex_name_attr =
+	__ATTR(sysplex_name, 0644, sysplex_name_show, sysplex_name_store);
+
+static ssize_t system_type_show(struct kobject *kobj,
+				struct kobj_attribute *attr, char *page)
+{
+	int rc;
+
+	mutex_lock(&sclp_cpi_mutex);
+	rc = snprintf(page, PAGE_SIZE, "%s\n", system_type);
+	mutex_unlock(&sclp_cpi_mutex);
+	return rc;
+}
+
+static ssize_t system_type_store(struct kobject *kobj,
+				 struct kobj_attribute *attr,
+				 const char *buf,
+	size_t len)
+{
+	int rc;
+
+	rc = check_string("system_type", buf);
+	if (rc)
+		return rc;
+
+	mutex_lock(&sclp_cpi_mutex);
+	set_string(system_type, buf);
+	mutex_unlock(&sclp_cpi_mutex);
+
+	return len;
+}
+
+static struct kobj_attribute system_type_attr =
+	__ATTR(system_type, 0644, system_type_show, system_type_store);
+
+static ssize_t system_level_show(struct kobject *kobj,
+				 struct kobj_attribute *attr, char *page)
+{
+	unsigned long long level;
+
+	mutex_lock(&sclp_cpi_mutex);
+	level = system_level;
+	mutex_unlock(&sclp_cpi_mutex);
+	return snprintf(page, PAGE_SIZE, "%#018llx\n", level);
+}
+
+static ssize_t system_level_store(struct kobject *kobj,
+				  struct kobj_attribute *attr,
+				  const char *buf,
+	size_t len)
+{
+	unsigned long long level;
+	char *endp;
+
+	level = simple_strtoull(buf, &endp, 16);
+
+	if (endp == buf)
+		return -EINVAL;
+	if (*endp == '\n')
+		endp++;
+	if (*endp)
+		return -EINVAL;
+
+	mutex_lock(&sclp_cpi_mutex);
+	system_level = level;
+	mutex_unlock(&sclp_cpi_mutex);
+	return len;
+}
+
+static struct kobj_attribute system_level_attr =
+	__ATTR(system_level, 0644, system_level_show, system_level_store);
+
+static ssize_t set_store(struct kobject *kobj,
+			 struct kobj_attribute *attr,
+			 const char *buf, size_t len)
+{
+	int rc;
+
+	mutex_lock(&sclp_cpi_mutex);
+	rc = cpi_req();
+	mutex_unlock(&sclp_cpi_mutex);
+	if (rc)
+		return rc;
+
+	return len;
+}
+
+static struct kobj_attribute set_attr = __ATTR(set, 0200, NULL, set_store);
+
+static struct attribute *cpi_attrs[] = {
+	&system_name_attr.attr,
+	&sysplex_name_attr.attr,
+	&system_type_attr.attr,
+	&system_level_attr.attr,
+	&set_attr.attr,
+	NULL,
+};
+
+static struct attribute_group cpi_attr_group = {
+	.attrs = cpi_attrs,
+};
+
+static struct kset *cpi_kset;
+
+int sclp_cpi_set_data(const char *system, const char *sysplex, const char *type,
+		      const u64 level)
+{
+	int rc;
+
+	rc = check_string("system_name", system);
+	if (rc)
+		return rc;
+	rc = check_string("sysplex_name", sysplex);
+	if (rc)
+		return rc;
+	rc = check_string("system_type", type);
+	if (rc)
+		return rc;
+
+	mutex_lock(&sclp_cpi_mutex);
+	set_string(system_name, system);
+	set_string(sysplex_name, sysplex);
+	set_string(system_type, type);
+	system_level = level;
+
+	rc = cpi_req();
+	mutex_unlock(&sclp_cpi_mutex);
+
+	return rc;
+}
+EXPORT_SYMBOL(sclp_cpi_set_data);
+
+static int __init cpi_init(void)
+{
+	int rc;
+
+	cpi_kset = kset_create_and_add("cpi", NULL, firmware_kobj);
+	if (!cpi_kset)
+		return -ENOMEM;
+
+	rc = sysfs_create_group(&cpi_kset->kobj, &cpi_attr_group);
+	if (rc)
+		kset_unregister(cpi_kset);
+
+	return rc;
+}
+
+__initcall(cpi_init);

drivers/s390/char/sclp_cpi_sys.h

@@ -0,0 +1,14 @@
+/*
+ *    SCLP control program identification sysfs interface
+ *
+ *    Copyright IBM Corp. 2007
+ *    Author(s): Michael Ernst <mernst@de.ibm.com>
+ */
+
+#ifndef __SCLP_CPI_SYS_H__
+#define __SCLP_CPI_SYS_H__
+
+int sclp_cpi_set_data(const char *system, const char *sysplex,
+		      const char *type, u64 level);
+
+#endif	 /* __SCLP_CPI_SYS_H__ */

drivers/s390/char/sclp_ctl.c

@@ -0,0 +1,144 @@
+/*
+ * IOCTL interface for SCLP
+ *
+ * Copyright IBM Corp. 2012
+ *
+ * Author: Michael Holzheu <holzheu@linux.vnet.ibm.com>
+ */
+
+#include <linux/compat.h>
+#include <linux/uaccess.h>
+#include <linux/miscdevice.h>
+#include <linux/gfp.h>
+#include <linux/module.h>
+#include <linux/ioctl.h>
+#include <linux/fs.h>
+#include <asm/compat.h>
+#include <asm/sclp_ctl.h>
+#include <asm/sclp.h>
+
+#include "sclp.h"
+
+/*
+ * Supported command words
+ */
+static unsigned int sclp_ctl_sccb_wlist[] = {
+	0x00400002,
+	0x00410002,
+};
+
+/*
+ * Check if command word is supported
+ */
+static int sclp_ctl_cmdw_supported(unsigned int cmdw)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(sclp_ctl_sccb_wlist); i++) {
+		if (cmdw == sclp_ctl_sccb_wlist[i])
+			return 1;
+	}
+	return 0;
+}
+
+static void __user *u64_to_uptr(u64 value)
+{
+	if (is_compat_task())
+		return compat_ptr(value);
+	else
+		return (void __user *)(unsigned long)value;
+}
+
+/*
+ * Start SCLP request
+ */
+static int sclp_ctl_ioctl_sccb(void __user *user_area)
+{
+	struct sclp_ctl_sccb ctl_sccb;
+	struct sccb_header *sccb;
+	int rc;
+
+	if (copy_from_user(&ctl_sccb, user_area, sizeof(ctl_sccb)))
+		return -EFAULT;
+	if (!sclp_ctl_cmdw_supported(ctl_sccb.cmdw))
+		return -EOPNOTSUPP;
+	sccb = (void *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+	if (!sccb)
+		return -ENOMEM;
+	if (copy_from_user(sccb, u64_to_uptr(ctl_sccb.sccb), sizeof(*sccb))) {
+		rc = -EFAULT;
+		goto out_free;
+	}
+	if (sccb->length > PAGE_SIZE || sccb->length < 8)
+		return -EINVAL;
+	if (copy_from_user(sccb, u64_to_uptr(ctl_sccb.sccb), sccb->length)) {
+		rc = -EFAULT;
+		goto out_free;
+	}
+	rc = sclp_sync_request(ctl_sccb.cmdw, sccb);
+	if (rc)
+		goto out_free;
+	if (copy_to_user(u64_to_uptr(ctl_sccb.sccb), sccb, sccb->length))
+		rc = -EFAULT;
+out_free:
+	free_page((unsigned long) sccb);
+	return rc;
+}
+
+/*
+ * SCLP SCCB ioctl function
+ */
+static long sclp_ctl_ioctl(struct file *filp, unsigned int cmd,
+			   unsigned long arg)
+{
+	void __user *argp;
+
+	if (is_compat_task())
+		argp = compat_ptr(arg);
+	else
+		argp = (void __user *) arg;
+	switch (cmd) {
+	case SCLP_CTL_SCCB:
+		return sclp_ctl_ioctl_sccb(argp);
+	default: /* unknown ioctl number */
+		return -ENOTTY;
+	}
+}
+
+/*
+ * File operations
+ */
+static const struct file_operations sclp_ctl_fops = {
+	.owner = THIS_MODULE,
+	.open = nonseekable_open,
+	.unlocked_ioctl = sclp_ctl_ioctl,
+	.compat_ioctl = sclp_ctl_ioctl,
+	.llseek = no_llseek,
+};
+
+/*
+ * Misc device definition
+ */
+static struct miscdevice sclp_ctl_device = {
+	.minor = MISC_DYNAMIC_MINOR,
+	.name = "sclp",
+	.fops = &sclp_ctl_fops,
+};
+
+/*
+ * Register sclp_ctl misc device
+ */
+static int __init sclp_ctl_init(void)
+{
+	return misc_register(&sclp_ctl_device);
+}
+module_init(sclp_ctl_init);
+
+/*
+ * Deregister sclp_ctl misc device
+ */
+static void __exit sclp_ctl_exit(void)
+{
+	misc_deregister(&sclp_ctl_device);
+}
+module_exit(sclp_ctl_exit);

drivers/s390/char/sclp_diag.h

@@ -0,0 +1,89 @@
+/*
+ *    Copyright IBM Corp. 2013
+ *    Author(s): Ralf Hoppe (rhoppe@de.ibm.com)
+ */
+
+#ifndef _SCLP_DIAG_H
+#define _SCLP_DIAG_H
+
+#include <linux/types.h>
+
+/* return codes for Diagnostic Test FTP Service, as indicated in member
+ * sclp_diag_ftp::ldflg
+ */
+#define SCLP_DIAG_FTP_OK	0x80U /* success */
+#define SCLP_DIAG_FTP_LDFAIL	0x01U /* load failed */
+#define SCLP_DIAG_FTP_LDNPERM	0x02U /* not allowed */
+#define SCLP_DIAG_FTP_LDRUNS	0x03U /* LD runs */
+#define SCLP_DIAG_FTP_LDNRUNS	0x04U /* LD does not run */
+
+#define SCLP_DIAG_FTP_XPCX	0x80 /* PCX communication code */
+#define SCLP_DIAG_FTP_ROUTE	4 /* routing code for new FTP service */
+
+/*
+ * length of Diagnostic Test FTP Service event buffer
+ */
+#define SCLP_DIAG_FTP_EVBUF_LEN				\
+	(offsetof(struct sclp_diag_evbuf, mdd) +	\
+	 sizeof(struct sclp_diag_ftp))
+
+/**
+ * struct sclp_diag_ftp - Diagnostic Test FTP Service model-dependent data
+ * @pcx: code for PCX communication (should be 0x80)
+ * @ldflg: load flag (see defines above)
+ * @cmd: FTP command
+ * @pgsize: page size (0 = 4kB, 1 = large page size)
+ * @srcflg: source flag
+ * @spare: reserved (zeroes)
+ * @offset: file offset
+ * @fsize: file size
+ * @length: buffer size resp. bytes transferred
+ * @failaddr: failing address
+ * @bufaddr: buffer address, virtual
+ * @asce: region or segment table designation
+ * @fident: file name (ASCII, zero-terminated)
+ */
+struct sclp_diag_ftp {
+	u8 pcx;
+	u8 ldflg;
+	u8 cmd;
+	u8 pgsize;
+	u8 srcflg;
+	u8 spare;
+	u64 offset;
+	u64 fsize;
+	u64 length;
+	u64 failaddr;
+	u64 bufaddr;
+	u64 asce;
+
+	u8 fident[256];
+} __packed;
+
+/**
+ * struct sclp_diag_evbuf - Diagnostic Test (ET7) Event Buffer
+ * @hdr: event buffer header
+ * @route: diagnostic route
+ * @mdd: model-dependent data (@route dependent)
+ */
+struct sclp_diag_evbuf {
+	struct evbuf_header hdr;
+	u16 route;
+
+	union {
+		struct sclp_diag_ftp ftp;
+	} mdd;
+} __packed;
+
+/**
+ * struct sclp_diag_sccb - Diagnostic Test (ET7) SCCB
+ * @hdr: SCCB header
+ * @evbuf: event buffer
+ */
+struct sclp_diag_sccb {
+
+	struct sccb_header hdr;
+	struct sclp_diag_evbuf evbuf;
+} __packed;
+
+#endif /* _SCLP_DIAG_H */

drivers/s390/char/sclp_early.c

@@ -0,0 +1,315 @@
+/*
+ * SCLP early driver
+ *
+ * Copyright IBM Corp. 2013
+ */
+
+#define KMSG_COMPONENT "sclp_early"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <asm/ctl_reg.h>
+#include <asm/sclp.h>
+#include <asm/ipl.h>
+#include "sclp_sdias.h"
+#include "sclp.h"
+
+#define SCLP_CMDW_READ_SCP_INFO		0x00020001
+#define SCLP_CMDW_READ_SCP_INFO_FORCED	0x00120001
+
+struct read_info_sccb {
+	struct	sccb_header header;	/* 0-7 */
+	u16	rnmax;			/* 8-9 */
+	u8	rnsize;			/* 10 */
+	u8	_reserved0[16 - 11];	/* 11-15 */
+	u16	ncpurl;			/* 16-17 */
+	u16	cpuoff;			/* 18-19 */
+	u8	_reserved7[24 - 20];	/* 20-23 */
+	u8	loadparm[8];		/* 24-31 */
+	u8	_reserved1[48 - 32];	/* 32-47 */
+	u64	facilities;		/* 48-55 */
+	u8	_reserved2a[76 - 56];	/* 56-75 */
+	u32	ibc;			/* 76-79 */
+	u8	_reserved2b[84 - 80];	/* 80-83 */
+	u8	fac84;			/* 84 */
+	u8	fac85;			/* 85 */
+	u8	_reserved3[91 - 86];	/* 86-90 */
+	u8	flags;			/* 91 */
+	u8	_reserved4[100 - 92];	/* 92-99 */
+	u32	rnsize2;		/* 100-103 */
+	u64	rnmax2;			/* 104-111 */
+	u8	_reserved5[120 - 112];	/* 112-119 */
+	u16	hcpua;			/* 120-121 */
+	u8	_reserved6[4096 - 122];	/* 122-4095 */
+} __packed __aligned(PAGE_SIZE);
+
+static char sccb_early[PAGE_SIZE] __aligned(PAGE_SIZE) __initdata;
+static unsigned int sclp_con_has_vt220 __initdata;
+static unsigned int sclp_con_has_linemode __initdata;
+static unsigned long sclp_hsa_size;
+static unsigned int sclp_max_cpu;
+static struct sclp_ipl_info sclp_ipl_info;
+static unsigned char sclp_siif;
+static u32 sclp_ibc;
+
+u64 sclp_facilities;
+u8 sclp_fac84;
+unsigned long long sclp_rzm;
+unsigned long long sclp_rnmax;
+
+static int __init sclp_cmd_sync_early(sclp_cmdw_t cmd, void *sccb)
+{
+	int rc;
+
+	__ctl_set_bit(0, 9);
+	rc = sclp_service_call(cmd, sccb);
+	if (rc)
+		goto out;
+	__load_psw_mask(PSW_DEFAULT_KEY | PSW_MASK_BASE | PSW_MASK_EA |
+			PSW_MASK_BA | PSW_MASK_EXT | PSW_MASK_WAIT);
+	local_irq_disable();
+out:
+	/* Contents of the sccb might have changed. */
+	barrier();
+	__ctl_clear_bit(0, 9);
+	return rc;
+}
+
+static int __init sclp_read_info_early(struct read_info_sccb *sccb)
+{
+	int rc, i;
+	sclp_cmdw_t commands[] = {SCLP_CMDW_READ_SCP_INFO_FORCED,
+				  SCLP_CMDW_READ_SCP_INFO};
+
+	for (i = 0; i < ARRAY_SIZE(commands); i++) {
+		do {
+			memset(sccb, 0, sizeof(*sccb));
+			sccb->header.length = sizeof(*sccb);
+			sccb->header.function_code = 0x80;
+			sccb->header.control_mask[2] = 0x80;
+			rc = sclp_cmd_sync_early(commands[i], sccb);
+		} while (rc == -EBUSY);
+
+		if (rc)
+			break;
+		if (sccb->header.response_code == 0x10)
+			return 0;
+		if (sccb->header.response_code != 0x1f0)
+			break;
+	}
+	return -EIO;
+}
+
+static void __init sclp_facilities_detect(struct read_info_sccb *sccb)
+{
+	struct sclp_cpu_entry *cpue;
+	u16 boot_cpu_address, cpu;
+
+	if (sclp_read_info_early(sccb))
+		return;
+
+	sclp_facilities = sccb->facilities;
+	sclp_fac84 = sccb->fac84;
+	if (sccb->fac85 & 0x02)
+		S390_lowcore.machine_flags |= MACHINE_FLAG_ESOP;
+	sclp_rnmax = sccb->rnmax ? sccb->rnmax : sccb->rnmax2;
+	sclp_rzm = sccb->rnsize ? sccb->rnsize : sccb->rnsize2;
+	sclp_rzm <<= 20;
+	sclp_ibc = sccb->ibc;
+
+	if (!sccb->hcpua) {
+		if (MACHINE_IS_VM)
+			sclp_max_cpu = 64;
+		else
+			sclp_max_cpu = sccb->ncpurl;
+	} else {
+		sclp_max_cpu = sccb->hcpua + 1;
+	}
+
+	boot_cpu_address = stap();
+	cpue = (void *)sccb + sccb->cpuoff;
+	for (cpu = 0; cpu < sccb->ncpurl; cpue++, cpu++) {
+		if (boot_cpu_address != cpue->address)
+			continue;
+		sclp_siif = cpue->siif;
+		break;
+	}
+
+	/* Save IPL information */
+	sclp_ipl_info.is_valid = 1;
+	if (sccb->flags & 0x2)
+		sclp_ipl_info.has_dump = 1;
+	memcpy(&sclp_ipl_info.loadparm, &sccb->loadparm, LOADPARM_LEN);
+}
+
+bool __init sclp_has_linemode(void)
+{
+	return !!sclp_con_has_linemode;
+}
+
+bool __init sclp_has_vt220(void)
+{
+	return !!sclp_con_has_vt220;
+}
+
+unsigned long long sclp_get_rnmax(void)
+{
+	return sclp_rnmax;
+}
+
+unsigned long long sclp_get_rzm(void)
+{
+	return sclp_rzm;
+}
+
+unsigned int sclp_get_max_cpu(void)
+{
+	return sclp_max_cpu;
+}
+
+int sclp_has_siif(void)
+{
+	return sclp_siif;
+}
+EXPORT_SYMBOL(sclp_has_siif);
+
+unsigned int sclp_get_ibc(void)
+{
+	return sclp_ibc;
+}
+EXPORT_SYMBOL(sclp_get_ibc);
+
+/*
+ * This function will be called after sclp_facilities_detect(), which gets
+ * called from early.c code. The sclp_facilities_detect() function retrieves
+ * and saves the IPL information.
+ */
+void __init sclp_get_ipl_info(struct sclp_ipl_info *info)
+{
+	*info = sclp_ipl_info;
+}
+
+static int __init sclp_cmd_early(sclp_cmdw_t cmd, void *sccb)
+{
+	int rc;
+
+	do {
+		rc = sclp_cmd_sync_early(cmd, sccb);
+	} while (rc == -EBUSY);
+
+	if (rc)
+		return -EIO;
+	if (((struct sccb_header *) sccb)->response_code != 0x0020)
+		return -EIO;
+	return 0;
+}
+
+static void __init sccb_init_eq_size(struct sdias_sccb *sccb)
+{
+	memset(sccb, 0, sizeof(*sccb));
+
+	sccb->hdr.length = sizeof(*sccb);
+	sccb->evbuf.hdr.length = sizeof(struct sdias_evbuf);
+	sccb->evbuf.hdr.type = EVTYP_SDIAS;
+	sccb->evbuf.event_qual = SDIAS_EQ_SIZE;
+	sccb->evbuf.data_id = SDIAS_DI_FCP_DUMP;
+	sccb->evbuf.event_id = 4712;
+	sccb->evbuf.dbs = 1;
+}
+
+static int __init sclp_set_event_mask(struct init_sccb *sccb,
+				      unsigned long receive_mask,
+				      unsigned long send_mask)
+{
+	memset(sccb, 0, sizeof(*sccb));
+	sccb->header.length = sizeof(*sccb);
+	sccb->mask_length = sizeof(sccb_mask_t);
+	sccb->receive_mask = receive_mask;
+	sccb->send_mask = send_mask;
+	return sclp_cmd_early(SCLP_CMDW_WRITE_EVENT_MASK, sccb);
+}
+
+static long __init sclp_hsa_size_init(struct sdias_sccb *sccb)
+{
+	sccb_init_eq_size(sccb);
+	if (sclp_cmd_early(SCLP_CMDW_WRITE_EVENT_DATA, sccb))
+		return -EIO;
+	if (sccb->evbuf.blk_cnt == 0)
+		return 0;
+	return (sccb->evbuf.blk_cnt - 1) * PAGE_SIZE;
+}
+
+static long __init sclp_hsa_copy_wait(struct sccb_header *sccb)
+{
+	memset(sccb, 0, PAGE_SIZE);
+	sccb->length = PAGE_SIZE;
+	if (sclp_cmd_early(SCLP_CMDW_READ_EVENT_DATA, sccb))
+		return -EIO;
+	if (((struct sdias_sccb *) sccb)->evbuf.blk_cnt == 0)
+		return 0;
+	return (((struct sdias_sccb *) sccb)->evbuf.blk_cnt - 1) * PAGE_SIZE;
+}
+
+unsigned long sclp_get_hsa_size(void)
+{
+	return sclp_hsa_size;
+}
+
+static void __init sclp_hsa_size_detect(void *sccb)
+{
+	long size;
+
+	/* First try synchronous interface (LPAR) */
+	if (sclp_set_event_mask(sccb, 0, 0x40000010))
+		return;
+	size = sclp_hsa_size_init(sccb);
+	if (size < 0)
+		return;
+	if (size != 0)
+		goto out;
+	/* Then try asynchronous interface (z/VM) */
+	if (sclp_set_event_mask(sccb, 0x00000010, 0x40000010))
+		return;
+	size = sclp_hsa_size_init(sccb);
+	if (size < 0)
+		return;
+	size = sclp_hsa_copy_wait(sccb);
+	if (size < 0)
+		return;
+out:
+	sclp_hsa_size = size;
+}
+
+static unsigned int __init sclp_con_check_linemode(struct init_sccb *sccb)
+{
+	if (!(sccb->sclp_send_mask & EVTYP_OPCMD_MASK))
+		return 0;
+	if (!(sccb->sclp_receive_mask & (EVTYP_MSG_MASK | EVTYP_PMSGCMD_MASK)))
+		return 0;
+	return 1;
+}
+
+static void __init sclp_console_detect(struct init_sccb *sccb)
+{
+	if (sccb->header.response_code != 0x20)
+		return;
+
+	if (sccb->sclp_send_mask & EVTYP_VT220MSG_MASK)
+		sclp_con_has_vt220 = 1;
+
+	if (sclp_con_check_linemode(sccb))
+		sclp_con_has_linemode = 1;
+}
+
+void __init sclp_early_detect(void)
+{
+	void *sccb = &sccb_early;
+
+	sclp_facilities_detect(sccb);
+	sclp_hsa_size_detect(sccb);
+
+	/* Turn off SCLP event notifications.  Also save remote masks in the
+	 * sccb.  These are sufficient to detect sclp console capabilities.
+	 */
+	sclp_set_event_mask(sccb, 0, 0);
+	sclp_console_detect(sccb);
+}

drivers/s390/char/sclp_ftp.c

@@ -0,0 +1,275 @@
+/*
+ *    SCLP Event Type (ET) 7 - Diagnostic Test FTP Services, useable on LPAR
+ *
+ *    Copyright IBM Corp. 2013
+ *    Author(s): Ralf Hoppe (rhoppe@de.ibm.com)
+ *
+ */
+
+#define KMSG_COMPONENT "hmcdrv"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/slab.h>
+#include <linux/io.h>
+#include <linux/wait.h>
+#include <linux/string.h>
+#include <linux/jiffies.h>
+#include <asm/sysinfo.h>
+#include <asm/ebcdic.h>
+
+#include "sclp.h"
+#include "sclp_diag.h"
+#include "sclp_ftp.h"
+
+static DECLARE_COMPLETION(sclp_ftp_rx_complete);
+static u8 sclp_ftp_ldflg;
+static u64 sclp_ftp_fsize;
+static u64 sclp_ftp_length;
+
+/**
+ * sclp_ftp_txcb() - Diagnostic Test FTP services SCLP command callback
+ */
+static void sclp_ftp_txcb(struct sclp_req *req, void *data)
+{
+	struct completion *completion = data;
+
+#ifdef DEBUG
+	pr_debug("SCLP (ET7) TX-IRQ, SCCB @ 0x%p: %*phN\n",
+		 req->sccb, 24, req->sccb);
+#endif
+	complete(completion);
+}
+
+/**
+ * sclp_ftp_rxcb() - Diagnostic Test FTP services receiver event callback
+ */
+static void sclp_ftp_rxcb(struct evbuf_header *evbuf)
+{
+	struct sclp_diag_evbuf *diag = (struct sclp_diag_evbuf *) evbuf;
+
+	/*
+	 * Check for Diagnostic Test FTP Service
+	 */
+	if (evbuf->type != EVTYP_DIAG_TEST ||
+	    diag->route != SCLP_DIAG_FTP_ROUTE ||
+	    diag->mdd.ftp.pcx != SCLP_DIAG_FTP_XPCX ||
+	    evbuf->length < SCLP_DIAG_FTP_EVBUF_LEN)
+		return;
+
+#ifdef DEBUG
+	pr_debug("SCLP (ET7) RX-IRQ, Event @ 0x%p: %*phN\n",
+		 evbuf, 24, evbuf);
+#endif
+
+	/*
+	 * Because the event buffer is located in a page which is owned
+	 * by the SCLP core, all data of interest must be copied. The
+	 * error indication is in 'sclp_ftp_ldflg'
+	 */
+	sclp_ftp_ldflg = diag->mdd.ftp.ldflg;
+	sclp_ftp_fsize = diag->mdd.ftp.fsize;
+	sclp_ftp_length = diag->mdd.ftp.length;
+
+	complete(&sclp_ftp_rx_complete);
+}
+
+/**
+ * sclp_ftp_et7() - start a Diagnostic Test FTP Service SCLP request
+ * @ftp: pointer to FTP descriptor
+ *
+ * Return: 0 on success, else a (negative) error code
+ */
+static int sclp_ftp_et7(const struct hmcdrv_ftp_cmdspec *ftp)
+{
+	struct completion completion;
+	struct sclp_diag_sccb *sccb;
+	struct sclp_req *req;
+	size_t len;
+	int rc;
+
+	req = kzalloc(sizeof(*req), GFP_KERNEL);
+	sccb = (void *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+	if (!req || !sccb) {
+		rc = -ENOMEM;
+		goto out_free;
+	}
+
+	sccb->hdr.length = SCLP_DIAG_FTP_EVBUF_LEN +
+		sizeof(struct sccb_header);
+	sccb->evbuf.hdr.type = EVTYP_DIAG_TEST;
+	sccb->evbuf.hdr.length = SCLP_DIAG_FTP_EVBUF_LEN;
+	sccb->evbuf.hdr.flags = 0; /* clear processed-buffer */
+	sccb->evbuf.route = SCLP_DIAG_FTP_ROUTE;
+	sccb->evbuf.mdd.ftp.pcx = SCLP_DIAG_FTP_XPCX;
+	sccb->evbuf.mdd.ftp.srcflg = 0;
+	sccb->evbuf.mdd.ftp.pgsize = 0;
+	sccb->evbuf.mdd.ftp.asce = _ASCE_REAL_SPACE;
+	sccb->evbuf.mdd.ftp.ldflg = SCLP_DIAG_FTP_LDFAIL;
+	sccb->evbuf.mdd.ftp.fsize = 0;
+	sccb->evbuf.mdd.ftp.cmd = ftp->id;
+	sccb->evbuf.mdd.ftp.offset = ftp->ofs;
+	sccb->evbuf.mdd.ftp.length = ftp->len;
+	sccb->evbuf.mdd.ftp.bufaddr = virt_to_phys(ftp->buf);
+
+	len = strlcpy(sccb->evbuf.mdd.ftp.fident, ftp->fname,
+		      HMCDRV_FTP_FIDENT_MAX);
+	if (len >= HMCDRV_FTP_FIDENT_MAX) {
+		rc = -EINVAL;
+		goto out_free;
+	}
+
+	req->command = SCLP_CMDW_WRITE_EVENT_DATA;
+	req->sccb = sccb;
+	req->status = SCLP_REQ_FILLED;
+	req->callback = sclp_ftp_txcb;
+	req->callback_data = &completion;
+
+	init_completion(&completion);
+
+	rc = sclp_add_request(req);
+	if (rc)
+		goto out_free;
+
+	/* Wait for end of ftp sclp command. */
+	wait_for_completion(&completion);
+
+#ifdef DEBUG
+	pr_debug("status of SCLP (ET7) request is 0x%04x (0x%02x)\n",
+		 sccb->hdr.response_code, sccb->evbuf.hdr.flags);
+#endif
+
+	/*
+	 * Check if sclp accepted the request. The data transfer runs
+	 * asynchronously and the completion is indicated with an
+	 * sclp ET7 event.
+	 */
+	if (req->status != SCLP_REQ_DONE ||
+	    (sccb->evbuf.hdr.flags & 0x80) == 0 || /* processed-buffer */
+	    (sccb->hdr.response_code & 0xffU) != 0x20U) {
+		rc = -EIO;
+	}
+
+out_free:
+	free_page((unsigned long) sccb);
+	kfree(req);
+	return rc;
+}
+
+/**
+ * sclp_ftp_cmd() - executes a HMC related SCLP Diagnose (ET7) FTP command
+ * @ftp: pointer to FTP command specification
+ * @fsize: return of file size (or NULL if undesirable)
+ *
+ * Attention: Notice that this function is not reentrant - so the caller
+ * must ensure locking.
+ *
+ * Return: number of bytes read/written or a (negative) error code
+ */
+ssize_t sclp_ftp_cmd(const struct hmcdrv_ftp_cmdspec *ftp, size_t *fsize)
+{
+	ssize_t len;
+#ifdef DEBUG
+	unsigned long start_jiffies;
+
+	pr_debug("starting SCLP (ET7), cmd %d for '%s' at %lld with %zd bytes\n",
+		 ftp->id, ftp->fname, (long long) ftp->ofs, ftp->len);
+	start_jiffies = jiffies;
+#endif
+
+	init_completion(&sclp_ftp_rx_complete);
+
+	/* Start ftp sclp command. */
+	len = sclp_ftp_et7(ftp);
+	if (len)
+		goto out_unlock;
+
+	/*
+	 * There is no way to cancel the sclp ET7 request, the code
+	 * needs to wait unconditionally until the transfer is complete.
+	 */
+	wait_for_completion(&sclp_ftp_rx_complete);
+
+#ifdef DEBUG
+	pr_debug("completed SCLP (ET7) request after %lu ms (all)\n",
+		 (jiffies - start_jiffies) * 1000 / HZ);
+	pr_debug("return code of SCLP (ET7) FTP Service is 0x%02x, with %lld/%lld bytes\n",
+		 sclp_ftp_ldflg, sclp_ftp_length, sclp_ftp_fsize);
+#endif
+
+	switch (sclp_ftp_ldflg) {
+	case SCLP_DIAG_FTP_OK:
+		len = sclp_ftp_length;
+		if (fsize)
+			*fsize = sclp_ftp_fsize;
+		break;
+	case SCLP_DIAG_FTP_LDNPERM:
+		len = -EPERM;
+		break;
+	case SCLP_DIAG_FTP_LDRUNS:
+		len = -EBUSY;
+		break;
+	case SCLP_DIAG_FTP_LDFAIL:
+		len = -ENOENT;
+		break;
+	default:
+		len = -EIO;
+		break;
+	}
+
+out_unlock:
+	return len;
+}
+
+/*
+ * ET7 event listener
+ */
+static struct sclp_register sclp_ftp_event = {
+	.send_mask = EVTYP_DIAG_TEST_MASK,    /* want tx events */
+	.receive_mask = EVTYP_DIAG_TEST_MASK, /* want rx events */
+	.receiver_fn = sclp_ftp_rxcb,	      /* async callback (rx) */
+	.state_change_fn = NULL,
+	.pm_event_fn = NULL,
+};
+
+/**
+ * sclp_ftp_startup() - startup of FTP services, when running on LPAR
+ */
+int sclp_ftp_startup(void)
+{
+#ifdef DEBUG
+	unsigned long info;
+#endif
+	int rc;
+
+	rc = sclp_register(&sclp_ftp_event);
+	if (rc)
+		return rc;
+
+#ifdef DEBUG
+	info = get_zeroed_page(GFP_KERNEL);
+
+	if (info != 0) {
+		struct sysinfo_2_2_2 *info222 = (struct sysinfo_2_2_2 *)info;
+
+		if (!stsi(info222, 2, 2, 2)) { /* get SYSIB 2.2.2 */
+			info222->name[sizeof(info222->name) - 1] = '\0';
+			EBCASC_500(info222->name, sizeof(info222->name) - 1);
+			pr_debug("SCLP (ET7) FTP Service working on LPAR %u (%s)\n",
+				 info222->lpar_number, info222->name);
+		}
+
+		free_page(info);
+	}
+#endif	/* DEBUG */
+	return 0;
+}
+
+/**
+ * sclp_ftp_shutdown() - shutdown of FTP services, when running on LPAR
+ */
+void sclp_ftp_shutdown(void)
+{
+	sclp_unregister(&sclp_ftp_event);
+}

drivers/s390/char/sclp_ftp.h

@@ -0,0 +1,21 @@
+/*
+ *    SCLP Event Type (ET) 7 - Diagnostic Test FTP Services, useable on LPAR
+ *
+ *    Notice that all functions exported here are not reentrant.
+ *    So usage should be exclusive, ensured by the caller (e.g. using a
+ *    mutex).
+ *
+ *    Copyright IBM Corp. 2013
+ *    Author(s): Ralf Hoppe (rhoppe@de.ibm.com)
+ */
+
+#ifndef __SCLP_FTP_H__
+#define __SCLP_FTP_H__
+
+#include "hmcdrv_ftp.h"
+
+int sclp_ftp_startup(void);
+void sclp_ftp_shutdown(void);
+ssize_t sclp_ftp_cmd(const struct hmcdrv_ftp_cmdspec *ftp, size_t *fsize);
+
+#endif	 /* __SCLP_FTP_H__ */

drivers/s390/char/sclp_ocf.c

@@ -0,0 +1,144 @@
+/*
+ *    SCLP OCF communication parameters sysfs interface
+ *
+ *    Copyright IBM Corp. 2011
+ *    Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#define KMSG_COMPONENT "sclp_ocf"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/stat.h>
+#include <linux/device.h>
+#include <linux/string.h>
+#include <linux/ctype.h>
+#include <linux/kmod.h>
+#include <linux/timer.h>
+#include <linux/err.h>
+#include <asm/ebcdic.h>
+#include <asm/sclp.h>
+
+#include "sclp.h"
+
+#define OCF_LENGTH_HMC_NETWORK 8UL
+#define OCF_LENGTH_CPC_NAME 8UL
+
+static char hmc_network[OCF_LENGTH_HMC_NETWORK + 1];
+static char cpc_name[OCF_LENGTH_CPC_NAME + 1];
+
+static DEFINE_SPINLOCK(sclp_ocf_lock);
+static struct work_struct sclp_ocf_change_work;
+
+static struct kset *ocf_kset;
+
+static void sclp_ocf_change_notify(struct work_struct *work)
+{
+	kobject_uevent(&ocf_kset->kobj, KOBJ_CHANGE);
+}
+
+/* Handler for OCF event. Look for the CPC image name. */
+static void sclp_ocf_handler(struct evbuf_header *evbuf)
+{
+	struct gds_vector *v;
+	struct gds_subvector *sv, *netid, *cpc;
+	size_t size;
+
+	/* Find the 0x9f00 block. */
+	v = sclp_find_gds_vector(evbuf + 1, (void *) evbuf + evbuf->length,
+				 0x9f00);
+	if (!v)
+		return;
+	/* Find the 0x9f22 block inside the 0x9f00 block. */
+	v = sclp_find_gds_vector(v + 1, (void *) v + v->length, 0x9f22);
+	if (!v)
+		return;
+	/* Find the 0x81 block inside the 0x9f22 block. */
+	sv = sclp_find_gds_subvector(v + 1, (void *) v + v->length, 0x81);
+	if (!sv)
+		return;
+	/* Find the 0x01 block inside the 0x81 block. */
+	netid = sclp_find_gds_subvector(sv + 1, (void *) sv + sv->length, 1);
+	/* Find the 0x02 block inside the 0x81 block. */
+	cpc = sclp_find_gds_subvector(sv + 1, (void *) sv + sv->length, 2);
+	/* Copy network name and cpc name. */
+	spin_lock(&sclp_ocf_lock);
+	if (netid) {
+		size = min(OCF_LENGTH_HMC_NETWORK, (size_t) netid->length);
+		memcpy(hmc_network, netid + 1, size);
+		EBCASC(hmc_network, size);
+		hmc_network[size] = 0;
+	}
+	if (cpc) {
+		size = min(OCF_LENGTH_CPC_NAME, (size_t) cpc->length);
+		memcpy(cpc_name, cpc + 1, size);
+		EBCASC(cpc_name, size);
+		cpc_name[size] = 0;
+	}
+	spin_unlock(&sclp_ocf_lock);
+	schedule_work(&sclp_ocf_change_work);
+}
+
+static struct sclp_register sclp_ocf_event = {
+	.receive_mask = EVTYP_OCF_MASK,
+	.receiver_fn = sclp_ocf_handler,
+};
+
+static ssize_t cpc_name_show(struct kobject *kobj,
+			     struct kobj_attribute *attr, char *page)
+{
+	int rc;
+
+	spin_lock_irq(&sclp_ocf_lock);
+	rc = snprintf(page, PAGE_SIZE, "%s\n", cpc_name);
+	spin_unlock_irq(&sclp_ocf_lock);
+	return rc;
+}
+
+static struct kobj_attribute cpc_name_attr =
+	__ATTR(cpc_name, 0444, cpc_name_show, NULL);
+
+static ssize_t hmc_network_show(struct kobject *kobj,
+				struct kobj_attribute *attr, char *page)
+{
+	int rc;
+
+	spin_lock_irq(&sclp_ocf_lock);
+	rc = snprintf(page, PAGE_SIZE, "%s\n", hmc_network);
+	spin_unlock_irq(&sclp_ocf_lock);
+	return rc;
+}
+
+static struct kobj_attribute hmc_network_attr =
+	__ATTR(hmc_network, 0444, hmc_network_show, NULL);
+
+static struct attribute *ocf_attrs[] = {
+	&cpc_name_attr.attr,
+	&hmc_network_attr.attr,
+	NULL,
+};
+
+static struct attribute_group ocf_attr_group = {
+	.attrs = ocf_attrs,
+};
+
+static int __init ocf_init(void)
+{
+	int rc;
+
+	INIT_WORK(&sclp_ocf_change_work, sclp_ocf_change_notify);
+	ocf_kset = kset_create_and_add("ocf", NULL, firmware_kobj);
+	if (!ocf_kset)
+		return -ENOMEM;
+
+	rc = sysfs_create_group(&ocf_kset->kobj, &ocf_attr_group);
+	if (rc) {
+		kset_unregister(ocf_kset);
+		return rc;
+	}
+
+	return sclp_register(&sclp_ocf_event);
+}
+
+device_initcall(ocf_init);

drivers/s390/char/sclp_quiesce.c

@@ -0,0 +1,84 @@
+/*
+ *     signal quiesce handler
+ *
+ *  Copyright IBM Corp. 1999, 2004
+ *  Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>
+ *             Peter Oberparleiter <peter.oberparleiter@de.ibm.com>
+ */
+
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/cpumask.h>
+#include <linux/smp.h>
+#include <linux/init.h>
+#include <linux/reboot.h>
+#include <linux/atomic.h>
+#include <asm/ptrace.h>
+#include <asm/smp.h>
+
+#include "sclp.h"
+
+static void (*old_machine_restart)(char *);
+static void (*old_machine_halt)(void);
+static void (*old_machine_power_off)(void);
+
+/* Shutdown handler. Signal completion of shutdown by loading special PSW. */
+static void do_machine_quiesce(void)
+{
+	psw_t quiesce_psw;
+
+	smp_send_stop();
+	quiesce_psw.mask =
+		PSW_MASK_BASE | PSW_MASK_EA | PSW_MASK_BA | PSW_MASK_WAIT;
+	quiesce_psw.addr = 0xfff;
+	__load_psw(quiesce_psw);
+}
+
+/* Handler for quiesce event. Start shutdown procedure. */
+static void sclp_quiesce_handler(struct evbuf_header *evbuf)
+{
+	if (_machine_restart != (void *) do_machine_quiesce) {
+		old_machine_restart = _machine_restart;
+		old_machine_halt = _machine_halt;
+		old_machine_power_off = _machine_power_off;
+		_machine_restart = (void *) do_machine_quiesce;
+		_machine_halt = do_machine_quiesce;
+		_machine_power_off = do_machine_quiesce;
+	}
+	ctrl_alt_del();
+}
+
+/* Undo machine restart/halt/power_off modification on resume */
+static void sclp_quiesce_pm_event(struct sclp_register *reg,
+				  enum sclp_pm_event sclp_pm_event)
+{
+	switch (sclp_pm_event) {
+	case SCLP_PM_EVENT_RESTORE:
+		if (old_machine_restart) {
+			_machine_restart = old_machine_restart;
+			_machine_halt = old_machine_halt;
+			_machine_power_off = old_machine_power_off;
+			old_machine_restart = NULL;
+			old_machine_halt = NULL;
+			old_machine_power_off = NULL;
+		}
+		break;
+	case SCLP_PM_EVENT_FREEZE:
+	case SCLP_PM_EVENT_THAW:
+		break;
+	}
+}
+
+static struct sclp_register sclp_quiesce_event = {
+	.receive_mask = EVTYP_SIGQUIESCE_MASK,
+	.receiver_fn = sclp_quiesce_handler,
+	.pm_event_fn = sclp_quiesce_pm_event
+};
+
+/* Initialize quiesce driver. */
+static int __init sclp_quiesce_init(void)
+{
+	return sclp_register(&sclp_quiesce_event);
+}
+
+module_init(sclp_quiesce_init);

drivers/s390/char/sclp_rw.c

@@ -0,0 +1,469 @@
+/*
+ * driver: reading from and writing to system console on S/390 via SCLP
+ *
+ * Copyright IBM Corp. 1999, 2009
+ *
+ * Author(s): Martin Peschke <mpeschke@de.ibm.com>
+ *	      Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#include <linux/kmod.h>
+#include <linux/types.h>
+#include <linux/err.h>
+#include <linux/string.h>
+#include <linux/spinlock.h>
+#include <linux/ctype.h>
+#include <asm/uaccess.h>
+
+#include "sclp.h"
+#include "sclp_rw.h"
+
+/*
+ * The room for the SCCB (only for writing) is not equal to a pages size
+ * (as it is specified as the maximum size in the SCLP documentation)
+ * because of the additional data structure described above.
+ */
+#define MAX_SCCB_ROOM (PAGE_SIZE - sizeof(struct sclp_buffer))
+
+static void sclp_rw_pm_event(struct sclp_register *reg,
+			     enum sclp_pm_event sclp_pm_event)
+{
+	sclp_console_pm_event(sclp_pm_event);
+}
+
+/* Event type structure for write message and write priority message */
+static struct sclp_register sclp_rw_event = {
+	.send_mask = EVTYP_MSG_MASK,
+	.pm_event_fn = sclp_rw_pm_event,
+};
+
+/*
+ * Setup a sclp write buffer. Gets a page as input (4K) and returns
+ * a pointer to a struct sclp_buffer structure that is located at the
+ * end of the input page. This reduces the buffer space by a few
+ * bytes but simplifies things.
+ */
+struct sclp_buffer *
+sclp_make_buffer(void *page, unsigned short columns, unsigned short htab)
+{
+	struct sclp_buffer *buffer;
+	struct write_sccb *sccb;
+
+	sccb = (struct write_sccb *) page;
+	/*
+	 * We keep the struct sclp_buffer structure at the end
+	 * of the sccb page.
+	 */
+	buffer = ((struct sclp_buffer *) ((addr_t) sccb + PAGE_SIZE)) - 1;
+	buffer->sccb = sccb;
+	buffer->retry_count = 0;
+	buffer->mto_number = 0;
+	buffer->mto_char_sum = 0;
+	buffer->current_line = NULL;
+	buffer->current_length = 0;
+	buffer->columns = columns;
+	buffer->htab = htab;
+
+	/* initialize sccb */
+	memset(sccb, 0, sizeof(struct write_sccb));
+	sccb->header.length = sizeof(struct write_sccb);
+	sccb->msg_buf.header.length = sizeof(struct msg_buf);
+	sccb->msg_buf.header.type = EVTYP_MSG;
+	sccb->msg_buf.mdb.header.length = sizeof(struct mdb);
+	sccb->msg_buf.mdb.header.type = 1;
+	sccb->msg_buf.mdb.header.tag = 0xD4C4C240;	/* ebcdic "MDB " */
+	sccb->msg_buf.mdb.header.revision_code = 1;
+	sccb->msg_buf.mdb.go.length = sizeof(struct go);
+	sccb->msg_buf.mdb.go.type = 1;
+
+	return buffer;
+}
+
+/*
+ * Return a pointer to the original page that has been used to create
+ * the buffer.
+ */
+void *
+sclp_unmake_buffer(struct sclp_buffer *buffer)
+{
+	return buffer->sccb;
+}
+
+/*
+ * Initialize a new Message Text Object (MTO) at the end of the provided buffer
+ * with enough room for max_len characters. Return 0 on success.
+ */
+static int
+sclp_initialize_mto(struct sclp_buffer *buffer, int max_len)
+{
+	struct write_sccb *sccb;
+	struct mto *mto;
+	int mto_size;
+
+	/* max size of new Message Text Object including message text  */
+	mto_size = sizeof(struct mto) + max_len;
+
+	/* check if current buffer sccb can contain the mto */
+	sccb = buffer->sccb;
+	if ((MAX_SCCB_ROOM - sccb->header.length) < mto_size)
+		return -ENOMEM;
+
+	/* find address of new message text object */
+	mto = (struct mto *)(((addr_t) sccb) + sccb->header.length);
+
+	/*
+	 * fill the new Message-Text Object,
+	 * starting behind the former last byte of the SCCB
+	 */
+	memset(mto, 0, sizeof(struct mto));
+	mto->length = sizeof(struct mto);
+	mto->type = 4;	/* message text object */
+	mto->line_type_flags = LNTPFLGS_ENDTEXT; /* end text */
+
+	/* set pointer to first byte after struct mto. */
+	buffer->current_line = (char *) (mto + 1);
+	buffer->current_length = 0;
+
+	return 0;
+}
+
+/*
+ * Finalize MTO initialized by sclp_initialize_mto(), updating the sizes of
+ * MTO, enclosing MDB, event buffer and SCCB.
+ */
+static void
+sclp_finalize_mto(struct sclp_buffer *buffer)
+{
+	struct write_sccb *sccb;
+	struct mto *mto;
+	int str_len, mto_size;
+
+	str_len = buffer->current_length;
+	buffer->current_line = NULL;
+	buffer->current_length = 0;
+
+	/* real size of new Message Text Object including message text	*/
+	mto_size = sizeof(struct mto) + str_len;
+
+	/* find address of new message text object */
+	sccb = buffer->sccb;
+	mto = (struct mto *)(((addr_t) sccb) + sccb->header.length);
+
+	/* set size of message text object */
+	mto->length = mto_size;
+
+	/*
+	 * update values of sizes
+	 * (SCCB, Event(Message) Buffer, Message Data Block)
+	 */
+	sccb->header.length += mto_size;
+	sccb->msg_buf.header.length += mto_size;
+	sccb->msg_buf.mdb.header.length += mto_size;
+
+	/*
+	 * count number of buffered messages (= number of Message Text
+	 * Objects) and number of buffered characters
+	 * for the SCCB currently used for buffering and at all
+	 */
+	buffer->mto_number++;
+	buffer->mto_char_sum += str_len;
+}
+
+/*
+ * processing of a message including escape characters,
+ * returns number of characters written to the output sccb
+ * ("processed" means that is not guaranteed that the character have already
+ *  been sent to the SCLP but that it will be done at least next time the SCLP
+ *  is not busy)
+ */
+int
+sclp_write(struct sclp_buffer *buffer, const unsigned char *msg, int count)
+{
+	int spaces, i_msg;
+	int rc;
+
+	/*
+	 * parse msg for escape sequences (\t,\v ...) and put formated
+	 * msg into an mto (created by sclp_initialize_mto).
+	 *
+	 * We have to do this work ourselfs because there is no support for
+	 * these characters on the native machine and only partial support
+	 * under VM (Why does VM interpret \n but the native machine doesn't ?)
+	 *
+	 * Depending on i/o-control setting the message is always written
+	 * immediately or we wait for a final new line maybe coming with the
+	 * next message. Besides we avoid a buffer overrun by writing its
+	 * content.
+	 *
+	 * RESTRICTIONS:
+	 *
+	 * \r and \b work within one line because we are not able to modify
+	 * previous output that have already been accepted by the SCLP.
+	 *
+	 * \t combined with following \r is not correctly represented because
+	 * \t is expanded to some spaces but \r does not know about a
+	 * previous \t and decreases the current position by one column.
+	 * This is in order to a slim and quick implementation.
+	 */
+	for (i_msg = 0; i_msg < count; i_msg++) {
+		switch (msg[i_msg]) {
+		case '\n':	/* new line, line feed (ASCII)	*/
+			/* check if new mto needs to be created */
+			if (buffer->current_line == NULL) {
+				rc = sclp_initialize_mto(buffer, 0);
+				if (rc)
+					return i_msg;
+			}
+			sclp_finalize_mto(buffer);
+			break;
+		case '\a':	/* bell, one for several times	*/
+			/* set SCLP sound alarm bit in General Object */
+			buffer->sccb->msg_buf.mdb.go.general_msg_flags |=
+				GNRLMSGFLGS_SNDALRM;
+			break;
+		case '\t':	/* horizontal tabulator	 */
+			/* check if new mto needs to be created */
+			if (buffer->current_line == NULL) {
+				rc = sclp_initialize_mto(buffer,
+							 buffer->columns);
+				if (rc)
+					return i_msg;
+			}
+			/* "go to (next htab-boundary + 1, same line)" */
+			do {
+				if (buffer->current_length >= buffer->columns)
+					break;
+				/* ok, add a blank */
+				*buffer->current_line++ = 0x40;
+				buffer->current_length++;
+			} while (buffer->current_length % buffer->htab);
+			break;
+		case '\f':	/* form feed  */
+		case '\v':	/* vertical tabulator  */
+			/* "go to (actual column, actual line + 1)" */
+			/* = new line, leading spaces */
+			if (buffer->current_line != NULL) {
+				spaces = buffer->current_length;
+				sclp_finalize_mto(buffer);
+				rc = sclp_initialize_mto(buffer,
+							 buffer->columns);
+				if (rc)
+					return i_msg;
+				memset(buffer->current_line, 0x40, spaces);
+				buffer->current_line += spaces;
+				buffer->current_length = spaces;
+			} else {
+				/* one an empty line this is the same as \n */
+				rc = sclp_initialize_mto(buffer,
+							 buffer->columns);
+				if (rc)
+					return i_msg;
+				sclp_finalize_mto(buffer);
+			}
+			break;
+		case '\b':	/* backspace  */
+			/* "go to (actual column - 1, actual line)" */
+			/* decrement counter indicating position, */
+			/* do not remove last character */
+			if (buffer->current_line != NULL &&
+			    buffer->current_length > 0) {
+				buffer->current_length--;
+				buffer->current_line--;
+			}
+			break;
+		case 0x00:	/* end of string  */
+			/* transfer current line to SCCB */
+			if (buffer->current_line != NULL)
+				sclp_finalize_mto(buffer);
+			/* skip the rest of the message including the 0 byte */
+			i_msg = count - 1;
+			break;
+		default:	/* no escape character	*/
+			/* do not output unprintable characters */
+			if (!isprint(msg[i_msg]))
+				break;
+			/* check if new mto needs to be created */
+			if (buffer->current_line == NULL) {
+				rc = sclp_initialize_mto(buffer,
+							 buffer->columns);
+				if (rc)
+					return i_msg;
+			}
+			*buffer->current_line++ = sclp_ascebc(msg[i_msg]);
+			buffer->current_length++;
+			break;
+		}
+		/* check if current mto is full */
+		if (buffer->current_line != NULL &&
+		    buffer->current_length >= buffer->columns)
+			sclp_finalize_mto(buffer);
+	}
+
+	/* return number of processed characters */
+	return i_msg;
+}
+
+/*
+ * Return the number of free bytes in the sccb
+ */
+int
+sclp_buffer_space(struct sclp_buffer *buffer)
+{
+	int count;
+
+	count = MAX_SCCB_ROOM - buffer->sccb->header.length;
+	if (buffer->current_line != NULL)
+		count -= sizeof(struct mto) + buffer->current_length;
+	return count;
+}
+
+/*
+ * Return number of characters in buffer
+ */
+int
+sclp_chars_in_buffer(struct sclp_buffer *buffer)
+{
+	int count;
+
+	count = buffer->mto_char_sum;
+	if (buffer->current_line != NULL)
+		count += buffer->current_length;
+	return count;
+}
+
+/*
+ * sets or provides some values that influence the drivers behaviour
+ */
+void
+sclp_set_columns(struct sclp_buffer *buffer, unsigned short columns)
+{
+	buffer->columns = columns;
+	if (buffer->current_line != NULL &&
+	    buffer->current_length > buffer->columns)
+		sclp_finalize_mto(buffer);
+}
+
+void
+sclp_set_htab(struct sclp_buffer *buffer, unsigned short htab)
+{
+	buffer->htab = htab;
+}
+
+/*
+ * called by sclp_console_init and/or sclp_tty_init
+ */
+int
+sclp_rw_init(void)
+{
+	static int init_done = 0;
+	int rc;
+
+	if (init_done)
+		return 0;
+
+	rc = sclp_register(&sclp_rw_event);
+	if (rc == 0)
+		init_done = 1;
+	return rc;
+}
+
+#define SCLP_BUFFER_MAX_RETRY		1
+
+/*
+ * second half of Write Event Data-function that has to be done after
+ * interruption indicating completion of Service Call.
+ */
+static void
+sclp_writedata_callback(struct sclp_req *request, void *data)
+{
+	int rc;
+	struct sclp_buffer *buffer;
+	struct write_sccb *sccb;
+
+	buffer = (struct sclp_buffer *) data;
+	sccb = buffer->sccb;
+
+	if (request->status == SCLP_REQ_FAILED) {
+		if (buffer->callback != NULL)
+			buffer->callback(buffer, -EIO);
+		return;
+	}
+	/* check SCLP response code and choose suitable action	*/
+	switch (sccb->header.response_code) {
+	case 0x0020 :
+		/* Normal completion, buffer processed, message(s) sent */
+		rc = 0;
+		break;
+
+	case 0x0340: /* Contained SCLP equipment check */
+		if (++buffer->retry_count > SCLP_BUFFER_MAX_RETRY) {
+			rc = -EIO;
+			break;
+		}
+		/* remove processed buffers and requeue rest */
+		if (sclp_remove_processed((struct sccb_header *) sccb) > 0) {
+			/* not all buffers were processed */
+			sccb->header.response_code = 0x0000;
+			buffer->request.status = SCLP_REQ_FILLED;
+			rc = sclp_add_request(request);
+			if (rc == 0)
+				return;
+		} else
+			rc = 0;
+		break;
+
+	case 0x0040: /* SCLP equipment check */
+	case 0x05f0: /* Target resource in improper state */
+		if (++buffer->retry_count > SCLP_BUFFER_MAX_RETRY) {
+			rc = -EIO;
+			break;
+		}
+		/* retry request */
+		sccb->header.response_code = 0x0000;
+		buffer->request.status = SCLP_REQ_FILLED;
+		rc = sclp_add_request(request);
+		if (rc == 0)
+			return;
+		break;
+	default:
+		if (sccb->header.response_code == 0x71f0)
+			rc = -ENOMEM;
+		else
+			rc = -EINVAL;
+		break;
+	}
+	if (buffer->callback != NULL)
+		buffer->callback(buffer, rc);
+}
+
+/*
+ * Setup the request structure in the struct sclp_buffer to do SCLP Write
+ * Event Data and pass the request to the core SCLP loop. Return zero on
+ * success, non-zero otherwise.
+ */
+int
+sclp_emit_buffer(struct sclp_buffer *buffer,
+		 void (*callback)(struct sclp_buffer *, int))
+{
+	struct write_sccb *sccb;
+
+	/* add current line if there is one */
+	if (buffer->current_line != NULL)
+		sclp_finalize_mto(buffer);
+
+	/* Are there messages in the output buffer ? */
+	if (buffer->mto_number == 0)
+		return -EIO;
+
+	sccb = buffer->sccb;
+	/* Use normal write message */
+	sccb->msg_buf.header.type = EVTYP_MSG;
+
+	buffer->request.command = SCLP_CMDW_WRITE_EVENT_DATA;
+	buffer->request.status = SCLP_REQ_FILLED;
+	buffer->request.callback = sclp_writedata_callback;
+	buffer->request.callback_data = buffer;
+	buffer->request.sccb = sccb;
+	buffer->callback = callback;
+	return sclp_add_request(&buffer->request);
+}

drivers/s390/char/sclp_rw.h

@@ -0,0 +1,101 @@
+/*
+ * interface to the SCLP-read/write driver
+ *
+ * Copyright IBM Corporation 1999, 2009
+ *
+ * Author(s): Martin Peschke <mpeschke@de.ibm.com>
+ *	      Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#ifndef __SCLP_RW_H__
+#define __SCLP_RW_H__
+
+#include <linux/list.h>
+
+struct mto {
+	u16 length;
+	u16 type;
+	u16 line_type_flags;
+	u8  alarm_control;
+	u8  _reserved[3];
+} __attribute__((packed));
+
+struct go {
+	u16 length;
+	u16 type;
+	u32 domid;
+	u8  hhmmss_time[8];
+	u8  th_time[3];
+	u8  reserved_0;
+	u8  dddyyyy_date[7];
+	u8  _reserved_1;
+	u16 general_msg_flags;
+	u8  _reserved_2[10];
+	u8  originating_system_name[8];
+	u8  job_guest_name[8];
+} __attribute__((packed));
+
+struct mdb_header {
+	u16 length;
+	u16 type;
+	u32 tag;
+	u32 revision_code;
+} __attribute__((packed));
+
+struct mdb {
+	struct mdb_header header;
+	struct go go;
+} __attribute__((packed));
+
+struct msg_buf {
+	struct evbuf_header header;
+	struct mdb mdb;
+} __attribute__((packed));
+
+struct write_sccb {
+	struct sccb_header header;
+	struct msg_buf msg_buf;
+} __attribute__((packed));
+
+/* The number of empty mto buffers that can be contained in a single sccb. */
+#define NR_EMPTY_MTO_PER_SCCB ((PAGE_SIZE - sizeof(struct sclp_buffer) - \
+			sizeof(struct write_sccb)) / sizeof(struct mto))
+
+/*
+ * data structure for information about list of SCCBs (only for writing),
+ * will be located at the end of a SCCBs page
+ */
+struct sclp_buffer {
+	struct list_head list;		/* list_head for sccb_info chain */
+	struct sclp_req request;
+	struct write_sccb *sccb;
+	char *current_line;
+	int current_length;
+	int retry_count;
+	/* output format settings */
+	unsigned short columns;
+	unsigned short htab;
+	/* statistics about this buffer */
+	unsigned int mto_char_sum;	/* # chars in sccb */
+	unsigned int mto_number;	/* # mtos in sccb */
+	/* Callback that is called after reaching final status. */
+	void (*callback)(struct sclp_buffer *, int);
+};
+
+int sclp_rw_init(void);
+struct sclp_buffer *sclp_make_buffer(void *, unsigned short, unsigned short);
+void *sclp_unmake_buffer(struct sclp_buffer *);
+int sclp_buffer_space(struct sclp_buffer *);
+int sclp_write(struct sclp_buffer *buffer, const unsigned char *, int);
+int sclp_emit_buffer(struct sclp_buffer *,void (*)(struct sclp_buffer *,int));
+void sclp_set_columns(struct sclp_buffer *, unsigned short);
+void sclp_set_htab(struct sclp_buffer *, unsigned short);
+int sclp_chars_in_buffer(struct sclp_buffer *);
+
+#ifdef CONFIG_SCLP_CONSOLE
+void sclp_console_pm_event(enum sclp_pm_event sclp_pm_event);
+#else
+static inline void sclp_console_pm_event(enum sclp_pm_event sclp_pm_event) { }
+#endif
+
+#endif	/* __SCLP_RW_H__ */

drivers/s390/char/sclp_sdias.c

@@ -0,0 +1,281 @@
+/*
+ * SCLP "store data in absolute storage"
+ *
+ * Copyright IBM Corp. 2003, 2013
+ * Author(s): Michael Holzheu
+ */
+
+#define KMSG_COMPONENT "sclp_sdias"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/completion.h>
+#include <linux/sched.h>
+#include <asm/sclp.h>
+#include <asm/debug.h>
+#include <asm/ipl.h>
+
+#include "sclp_sdias.h"
+#include "sclp.h"
+#include "sclp_rw.h"
+
+#define TRACE(x...) debug_sprintf_event(sdias_dbf, 1, x)
+
+#define SDIAS_RETRIES 300
+#define SDIAS_SLEEP_TICKS 50
+
+static struct debug_info *sdias_dbf;
+
+static struct sclp_register sclp_sdias_register = {
+	.send_mask = EVTYP_SDIAS_MASK,
+};
+
+static struct sdias_sccb sccb __attribute__((aligned(4096)));
+static struct sdias_evbuf sdias_evbuf;
+
+static DECLARE_COMPLETION(evbuf_accepted);
+static DECLARE_COMPLETION(evbuf_done);
+static DEFINE_MUTEX(sdias_mutex);
+
+/*
+ * Called by SCLP base when read event data has been completed (async mode only)
+ */
+static void sclp_sdias_receiver_fn(struct evbuf_header *evbuf)
+{
+	memcpy(&sdias_evbuf, evbuf,
+	       min_t(unsigned long, sizeof(sdias_evbuf), evbuf->length));
+	complete(&evbuf_done);
+	TRACE("sclp_sdias_receiver_fn done\n");
+}
+
+/*
+ * Called by SCLP base when sdias event has been accepted
+ */
+static void sdias_callback(struct sclp_req *request, void *data)
+{
+	complete(&evbuf_accepted);
+	TRACE("callback done\n");
+}
+
+static int sdias_sclp_send(struct sclp_req *req)
+{
+	int retries;
+	int rc;
+
+	for (retries = SDIAS_RETRIES; retries; retries--) {
+		TRACE("add request\n");
+		rc = sclp_add_request(req);
+		if (rc) {
+			/* not initiated, wait some time and retry */
+			set_current_state(TASK_INTERRUPTIBLE);
+			TRACE("add request failed: rc = %i\n",rc);
+			schedule_timeout(SDIAS_SLEEP_TICKS);
+			continue;
+		}
+		/* initiated, wait for completion of service call */
+		wait_for_completion(&evbuf_accepted);
+		if (req->status == SCLP_REQ_FAILED) {
+			TRACE("sclp request failed\n");
+			continue;
+		}
+		/* if not accepted, retry */
+		if (!(sccb.evbuf.hdr.flags & 0x80)) {
+			TRACE("sclp request failed: flags=%x\n",
+			      sccb.evbuf.hdr.flags);
+			continue;
+		}
+		/*
+		 * for the sync interface the response is in the initial sccb
+		 */
+		if (!sclp_sdias_register.receiver_fn) {
+			memcpy(&sdias_evbuf, &sccb.evbuf, sizeof(sdias_evbuf));
+			TRACE("sync request done\n");
+			return 0;
+		}
+		/* otherwise we wait for completion */
+		wait_for_completion(&evbuf_done);
+		TRACE("request done\n");
+		return 0;
+	}
+	return -EIO;
+}
+
+/*
+ * Get number of blocks (4K) available in the HSA
+ */
+int sclp_sdias_blk_count(void)
+{
+	struct sclp_req request;
+	int rc;
+
+	mutex_lock(&sdias_mutex);
+
+	memset(&sccb, 0, sizeof(sccb));
+	memset(&request, 0, sizeof(request));
+
+	sccb.hdr.length = sizeof(sccb);
+	sccb.evbuf.hdr.length = sizeof(struct sdias_evbuf);
+	sccb.evbuf.hdr.type = EVTYP_SDIAS;
+	sccb.evbuf.event_qual = SDIAS_EQ_SIZE;
+	sccb.evbuf.data_id = SDIAS_DI_FCP_DUMP;
+	sccb.evbuf.event_id = 4712;
+	sccb.evbuf.dbs = 1;
+
+	request.sccb = &sccb;
+	request.command = SCLP_CMDW_WRITE_EVENT_DATA;
+	request.status = SCLP_REQ_FILLED;
+	request.callback = sdias_callback;
+
+	rc = sdias_sclp_send(&request);
+	if (rc) {
+		pr_err("sclp_send failed for get_nr_blocks\n");
+		goto out;
+	}
+	if (sccb.hdr.response_code != 0x0020) {
+		TRACE("send failed: %x\n", sccb.hdr.response_code);
+		rc = -EIO;
+		goto out;
+	}
+
+	switch (sdias_evbuf.event_status) {
+		case 0:
+			rc = sdias_evbuf.blk_cnt;
+			break;
+		default:
+			pr_err("SCLP error: %x\n", sdias_evbuf.event_status);
+			rc = -EIO;
+			goto out;
+	}
+	TRACE("%i blocks\n", rc);
+out:
+	mutex_unlock(&sdias_mutex);
+	return rc;
+}
+
+/*
+ * Copy from HSA to absolute storage (not reentrant):
+ *
+ * @dest     : Address of buffer where data should be copied
+ * @start_blk: Start Block (beginning with 1)
+ * @nr_blks  : Number of 4K blocks to copy
+ *
+ * Return Value: 0 : Requested 'number' of blocks of data copied
+ *		 <0: ERROR - negative event status
+ */
+int sclp_sdias_copy(void *dest, int start_blk, int nr_blks)
+{
+	struct sclp_req request;
+	int rc;
+
+	mutex_lock(&sdias_mutex);
+
+	memset(&sccb, 0, sizeof(sccb));
+	memset(&request, 0, sizeof(request));
+
+	sccb.hdr.length = sizeof(sccb);
+	sccb.evbuf.hdr.length = sizeof(struct sdias_evbuf);
+	sccb.evbuf.hdr.type = EVTYP_SDIAS;
+	sccb.evbuf.hdr.flags = 0;
+	sccb.evbuf.event_qual = SDIAS_EQ_STORE_DATA;
+	sccb.evbuf.data_id = SDIAS_DI_FCP_DUMP;
+	sccb.evbuf.event_id = 4712;
+#ifdef CONFIG_64BIT
+	sccb.evbuf.asa_size = SDIAS_ASA_SIZE_64;
+#else
+	sccb.evbuf.asa_size = SDIAS_ASA_SIZE_32;
+#endif
+	sccb.evbuf.event_status = 0;
+	sccb.evbuf.blk_cnt = nr_blks;
+	sccb.evbuf.asa = (unsigned long)dest;
+	sccb.evbuf.fbn = start_blk;
+	sccb.evbuf.lbn = 0;
+	sccb.evbuf.dbs = 1;
+
+	request.sccb	 = &sccb;
+	request.command  = SCLP_CMDW_WRITE_EVENT_DATA;
+	request.status	 = SCLP_REQ_FILLED;
+	request.callback = sdias_callback;
+
+	rc = sdias_sclp_send(&request);
+	if (rc) {
+		pr_err("sclp_send failed: %x\n", rc);
+		goto out;
+	}
+	if (sccb.hdr.response_code != 0x0020) {
+		TRACE("copy failed: %x\n", sccb.hdr.response_code);
+		rc = -EIO;
+		goto out;
+	}
+
+	switch (sdias_evbuf.event_status) {
+	case SDIAS_EVSTATE_ALL_STORED:
+		TRACE("all stored\n");
+		break;
+	case SDIAS_EVSTATE_PART_STORED:
+		TRACE("part stored: %i\n", sdias_evbuf.blk_cnt);
+		break;
+	case SDIAS_EVSTATE_NO_DATA:
+		TRACE("no data\n");
+		/* fall through */
+	default:
+		pr_err("Error from SCLP while copying hsa. Event status = %x\n",
+		       sdias_evbuf.event_status);
+		rc = -EIO;
+	}
+out:
+	mutex_unlock(&sdias_mutex);
+	return rc;
+}
+
+static int __init sclp_sdias_register_check(void)
+{
+	int rc;
+
+	rc = sclp_register(&sclp_sdias_register);
+	if (rc)
+		return rc;
+	if (sclp_sdias_blk_count() == 0) {
+		sclp_unregister(&sclp_sdias_register);
+		return -ENODEV;
+	}
+	return 0;
+}
+
+static int __init sclp_sdias_init_sync(void)
+{
+	TRACE("Try synchronous mode\n");
+	sclp_sdias_register.receive_mask = 0;
+	sclp_sdias_register.receiver_fn = NULL;
+	return sclp_sdias_register_check();
+}
+
+static int __init sclp_sdias_init_async(void)
+{
+	TRACE("Try asynchronous mode\n");
+	sclp_sdias_register.receive_mask = EVTYP_SDIAS_MASK;
+	sclp_sdias_register.receiver_fn = sclp_sdias_receiver_fn;
+	return sclp_sdias_register_check();
+}
+
+int __init sclp_sdias_init(void)
+{
+	if (ipl_info.type != IPL_TYPE_FCP_DUMP)
+		return 0;
+	sdias_dbf = debug_register("dump_sdias", 4, 1, 4 * sizeof(long));
+	debug_register_view(sdias_dbf, &debug_sprintf_view);
+	debug_set_level(sdias_dbf, 6);
+	if (sclp_sdias_init_sync() == 0)
+		goto out;
+	if (sclp_sdias_init_async() == 0)
+		goto out;
+	TRACE("init failed\n");
+	return -ENODEV;
+out:
+	TRACE("init done\n");
+	return 0;
+}
+
+void __exit sclp_sdias_exit(void)
+{
+	debug_unregister(sdias_dbf);
+	sclp_unregister(&sclp_sdias_register);
+}

drivers/s390/char/sclp_sdias.h

@@ -0,0 +1,46 @@
+/*
+ * SCLP "store data in absolute storage"
+ *
+ * Copyright IBM Corp. 2003, 2013
+ */
+
+#ifndef SCLP_SDIAS_H
+#define SCLP_SDIAS_H
+
+#include "sclp.h"
+
+#define SDIAS_EQ_STORE_DATA		0x0
+#define SDIAS_EQ_SIZE			0x1
+#define SDIAS_DI_FCP_DUMP		0x0
+#define SDIAS_ASA_SIZE_32		0x0
+#define SDIAS_ASA_SIZE_64		0x1
+#define SDIAS_EVSTATE_ALL_STORED	0x0
+#define SDIAS_EVSTATE_NO_DATA		0x3
+#define SDIAS_EVSTATE_PART_STORED	0x10
+
+struct sdias_evbuf {
+	struct	evbuf_header hdr;
+	u8	event_qual;
+	u8	data_id;
+	u64	reserved2;
+	u32	event_id;
+	u16	reserved3;
+	u8	asa_size;
+	u8	event_status;
+	u32	reserved4;
+	u32	blk_cnt;
+	u64	asa;
+	u32	reserved5;
+	u32	fbn;
+	u32	reserved6;
+	u32	lbn;
+	u16	reserved7;
+	u16	dbs;
+} __packed;
+
+struct sdias_sccb {
+	struct sccb_header	hdr;
+	struct sdias_evbuf	evbuf;
+} __packed;
+
+#endif /* SCLP_SDIAS_H */

drivers/s390/char/sclp_tty.c

@@ -0,0 +1,576 @@
+/*
+ *    SCLP line mode terminal driver.
+ *
+ *  S390 version
+ *    Copyright IBM Corp. 1999
+ *    Author(s): Martin Peschke <mpeschke@de.ibm.com>
+ *		 Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#include <linux/module.h>
+#include <linux/kmod.h>
+#include <linux/tty.h>
+#include <linux/tty_driver.h>
+#include <linux/tty_flip.h>
+#include <linux/err.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/gfp.h>
+#include <asm/uaccess.h>
+
+#include "ctrlchar.h"
+#include "sclp.h"
+#include "sclp_rw.h"
+#include "sclp_tty.h"
+
+/*
+ * size of a buffer that collects single characters coming in
+ * via sclp_tty_put_char()
+ */
+#define SCLP_TTY_BUF_SIZE 512
+
+/*
+ * There is exactly one SCLP terminal, so we can keep things simple
+ * and allocate all variables statically.
+ */
+
+/* Lock to guard over changes to global variables. */
+static spinlock_t sclp_tty_lock;
+/* List of free pages that can be used for console output buffering. */
+static struct list_head sclp_tty_pages;
+/* List of full struct sclp_buffer structures ready for output. */
+static struct list_head sclp_tty_outqueue;
+/* Counter how many buffers are emitted. */
+static int sclp_tty_buffer_count;
+/* Pointer to current console buffer. */
+static struct sclp_buffer *sclp_ttybuf;
+/* Timer for delayed output of console messages. */
+static struct timer_list sclp_tty_timer;
+
+static struct tty_port sclp_port;
+static unsigned char sclp_tty_chars[SCLP_TTY_BUF_SIZE];
+static unsigned short int sclp_tty_chars_count;
+
+struct tty_driver *sclp_tty_driver;
+
+static int sclp_tty_tolower;
+static int sclp_tty_columns = 80;
+
+#define SPACES_PER_TAB 8
+#define CASE_DELIMITER 0x6c /* to separate upper and lower case (% in EBCDIC) */
+
+/* This routine is called whenever we try to open a SCLP terminal. */
+static int
+sclp_tty_open(struct tty_struct *tty, struct file *filp)
+{
+	tty_port_tty_set(&sclp_port, tty);
+	tty->driver_data = NULL;
+	sclp_port.low_latency = 0;
+	return 0;
+}
+
+/* This routine is called when the SCLP terminal is closed. */
+static void
+sclp_tty_close(struct tty_struct *tty, struct file *filp)
+{
+	if (tty->count > 1)
+		return;
+	tty_port_tty_set(&sclp_port, NULL);
+}
+
+/*
+ * This routine returns the numbers of characters the tty driver
+ * will accept for queuing to be written.  This number is subject
+ * to change as output buffers get emptied, or if the output flow
+ * control is acted. This is not an exact number because not every
+ * character needs the same space in the sccb. The worst case is
+ * a string of newlines. Every newlines creates a new mto which
+ * needs 8 bytes.
+ */
+static int
+sclp_tty_write_room (struct tty_struct *tty)
+{
+	unsigned long flags;
+	struct list_head *l;
+	int count;
+
+	spin_lock_irqsave(&sclp_tty_lock, flags);
+	count = 0;
+	if (sclp_ttybuf != NULL)
+		count = sclp_buffer_space(sclp_ttybuf) / sizeof(struct mto);
+	list_for_each(l, &sclp_tty_pages)
+		count += NR_EMPTY_MTO_PER_SCCB;
+	spin_unlock_irqrestore(&sclp_tty_lock, flags);
+	return count;
+}
+
+static void
+sclp_ttybuf_callback(struct sclp_buffer *buffer, int rc)
+{
+	unsigned long flags;
+	void *page;
+
+	do {
+		page = sclp_unmake_buffer(buffer);
+		spin_lock_irqsave(&sclp_tty_lock, flags);
+		/* Remove buffer from outqueue */
+		list_del(&buffer->list);
+		sclp_tty_buffer_count--;
+		list_add_tail((struct list_head *) page, &sclp_tty_pages);
+		/* Check if there is a pending buffer on the out queue. */
+		buffer = NULL;
+		if (!list_empty(&sclp_tty_outqueue))
+			buffer = list_entry(sclp_tty_outqueue.next,
+					    struct sclp_buffer, list);
+		spin_unlock_irqrestore(&sclp_tty_lock, flags);
+	} while (buffer && sclp_emit_buffer(buffer, sclp_ttybuf_callback));
+
+	tty_port_tty_wakeup(&sclp_port);
+}
+
+static inline void
+__sclp_ttybuf_emit(struct sclp_buffer *buffer)
+{
+	unsigned long flags;
+	int count;
+	int rc;
+
+	spin_lock_irqsave(&sclp_tty_lock, flags);
+	list_add_tail(&buffer->list, &sclp_tty_outqueue);
+	count = sclp_tty_buffer_count++;
+	spin_unlock_irqrestore(&sclp_tty_lock, flags);
+	if (count)
+		return;
+	rc = sclp_emit_buffer(buffer, sclp_ttybuf_callback);
+	if (rc)
+		sclp_ttybuf_callback(buffer, rc);
+}
+
+/*
+ * When this routine is called from the timer then we flush the
+ * temporary write buffer.
+ */
+static void
+sclp_tty_timeout(unsigned long data)
+{
+	unsigned long flags;
+	struct sclp_buffer *buf;
+
+	spin_lock_irqsave(&sclp_tty_lock, flags);
+	buf = sclp_ttybuf;
+	sclp_ttybuf = NULL;
+	spin_unlock_irqrestore(&sclp_tty_lock, flags);
+
+	if (buf != NULL) {
+		__sclp_ttybuf_emit(buf);
+	}
+}
+
+/*
+ * Write a string to the sclp tty.
+ */
+static int sclp_tty_write_string(const unsigned char *str, int count, int may_fail)
+{
+	unsigned long flags;
+	void *page;
+	int written;
+	int overall_written;
+	struct sclp_buffer *buf;
+
+	if (count <= 0)
+		return 0;
+	overall_written = 0;
+	spin_lock_irqsave(&sclp_tty_lock, flags);
+	do {
+		/* Create a sclp output buffer if none exists yet */
+		if (sclp_ttybuf == NULL) {
+			while (list_empty(&sclp_tty_pages)) {
+				spin_unlock_irqrestore(&sclp_tty_lock, flags);
+				if (may_fail)
+					goto out;
+				else
+					sclp_sync_wait();
+				spin_lock_irqsave(&sclp_tty_lock, flags);
+			}
+			page = sclp_tty_pages.next;
+			list_del((struct list_head *) page);
+			sclp_ttybuf = sclp_make_buffer(page, sclp_tty_columns,
+						       SPACES_PER_TAB);
+		}
+		/* try to write the string to the current output buffer */
+		written = sclp_write(sclp_ttybuf, str, count);
+		overall_written += written;
+		if (written == count)
+			break;
+		/*
+		 * Not all characters could be written to the current
+		 * output buffer. Emit the buffer, create a new buffer
+		 * and then output the rest of the string.
+		 */
+		buf = sclp_ttybuf;
+		sclp_ttybuf = NULL;
+		spin_unlock_irqrestore(&sclp_tty_lock, flags);
+		__sclp_ttybuf_emit(buf);
+		spin_lock_irqsave(&sclp_tty_lock, flags);
+		str += written;
+		count -= written;
+	} while (count > 0);
+	/* Setup timer to output current console buffer after 1/10 second */
+	if (sclp_ttybuf && sclp_chars_in_buffer(sclp_ttybuf) &&
+	    !timer_pending(&sclp_tty_timer)) {
+		init_timer(&sclp_tty_timer);
+		sclp_tty_timer.function = sclp_tty_timeout;
+		sclp_tty_timer.data = 0UL;
+		sclp_tty_timer.expires = jiffies + HZ/10;
+		add_timer(&sclp_tty_timer);
+	}
+	spin_unlock_irqrestore(&sclp_tty_lock, flags);
+out:
+	return overall_written;
+}
+
+/*
+ * This routine is called by the kernel to write a series of characters to the
+ * tty device. The characters may come from user space or kernel space. This
+ * routine will return the number of characters actually accepted for writing.
+ */
+static int
+sclp_tty_write(struct tty_struct *tty, const unsigned char *buf, int count)
+{
+	if (sclp_tty_chars_count > 0) {
+		sclp_tty_write_string(sclp_tty_chars, sclp_tty_chars_count, 0);
+		sclp_tty_chars_count = 0;
+	}
+	return sclp_tty_write_string(buf, count, 1);
+}
+
+/*
+ * This routine is called by the kernel to write a single character to the tty
+ * device. If the kernel uses this routine, it must call the flush_chars()
+ * routine (if defined) when it is done stuffing characters into the driver.
+ *
+ * Characters provided to sclp_tty_put_char() are buffered by the SCLP driver.
+ * If the given character is a '\n' the contents of the SCLP write buffer
+ * - including previous characters from sclp_tty_put_char() and strings from
+ * sclp_write() without final '\n' - will be written.
+ */
+static int
+sclp_tty_put_char(struct tty_struct *tty, unsigned char ch)
+{
+	sclp_tty_chars[sclp_tty_chars_count++] = ch;
+	if (ch == '\n' || sclp_tty_chars_count >= SCLP_TTY_BUF_SIZE) {
+		sclp_tty_write_string(sclp_tty_chars, sclp_tty_chars_count, 0);
+		sclp_tty_chars_count = 0;
+	}
+	return 1;
+}
+
+/*
+ * This routine is called by the kernel after it has written a series of
+ * characters to the tty device using put_char().
+ */
+static void
+sclp_tty_flush_chars(struct tty_struct *tty)
+{
+	if (sclp_tty_chars_count > 0) {
+		sclp_tty_write_string(sclp_tty_chars, sclp_tty_chars_count, 0);
+		sclp_tty_chars_count = 0;
+	}
+}
+
+/*
+ * This routine returns the number of characters in the write buffer of the
+ * SCLP driver. The provided number includes all characters that are stored
+ * in the SCCB (will be written next time the SCLP is not busy) as well as
+ * characters in the write buffer (will not be written as long as there is a
+ * final line feed missing).
+ */
+static int
+sclp_tty_chars_in_buffer(struct tty_struct *tty)
+{
+	unsigned long flags;
+	struct list_head *l;
+	struct sclp_buffer *t;
+	int count;
+
+	spin_lock_irqsave(&sclp_tty_lock, flags);
+	count = 0;
+	if (sclp_ttybuf != NULL)
+		count = sclp_chars_in_buffer(sclp_ttybuf);
+	list_for_each(l, &sclp_tty_outqueue) {
+		t = list_entry(l, struct sclp_buffer, list);
+		count += sclp_chars_in_buffer(t);
+	}
+	spin_unlock_irqrestore(&sclp_tty_lock, flags);
+	return count;
+}
+
+/*
+ * removes all content from buffers of low level driver
+ */
+static void
+sclp_tty_flush_buffer(struct tty_struct *tty)
+{
+	if (sclp_tty_chars_count > 0) {
+		sclp_tty_write_string(sclp_tty_chars, sclp_tty_chars_count, 0);
+		sclp_tty_chars_count = 0;
+	}
+}
+
+/*
+ * push input to tty
+ */
+static void
+sclp_tty_input(unsigned char* buf, unsigned int count)
+{
+	struct tty_struct *tty = tty_port_tty_get(&sclp_port);
+	unsigned int cchar;
+
+	/*
+	 * If this tty driver is currently closed
+	 * then throw the received input away.
+	 */
+	if (tty == NULL)
+		return;
+	cchar = ctrlchar_handle(buf, count, tty);
+	switch (cchar & CTRLCHAR_MASK) {
+	case CTRLCHAR_SYSRQ:
+		break;
+	case CTRLCHAR_CTRL:
+		tty_insert_flip_char(&sclp_port, cchar, TTY_NORMAL);
+		tty_flip_buffer_push(&sclp_port);
+		break;
+	case CTRLCHAR_NONE:
+		/* send (normal) input to line discipline */
+		if (count < 2 ||
+		    (strncmp((const char *) buf + count - 2, "^n", 2) &&
+		     strncmp((const char *) buf + count - 2, "\252n", 2))) {
+			/* add the auto \n */
+			tty_insert_flip_string(&sclp_port, buf, count);
+			tty_insert_flip_char(&sclp_port, '\n', TTY_NORMAL);
+		} else
+			tty_insert_flip_string(&sclp_port, buf, count - 2);
+		tty_flip_buffer_push(&sclp_port);
+		break;
+	}
+	tty_kref_put(tty);
+}
+
+/*
+ * get a EBCDIC string in upper/lower case,
+ * find out characters in lower/upper case separated by a special character,
+ * modifiy original string,
+ * returns length of resulting string
+ */
+static int sclp_switch_cases(unsigned char *buf, int count)
+{
+	unsigned char *ip, *op;
+	int toggle;
+
+	/* initially changing case is off */
+	toggle = 0;
+	ip = op = buf;
+	while (count-- > 0) {
+		/* compare with special character */
+		if (*ip == CASE_DELIMITER) {
+			/* followed by another special character? */
+			if (count && ip[1] == CASE_DELIMITER) {
+				/*
+				 * ... then put a single copy of the special
+				 * character to the output string
+				 */
+				*op++ = *ip++;
+				count--;
+			} else
+				/*
+				 * ... special character follower by a normal
+				 * character toggles the case change behaviour
+				 */
+				toggle = ~toggle;
+			/* skip special character */
+			ip++;
+		} else
+			/* not the special character */
+			if (toggle)
+				/* but case switching is on */
+				if (sclp_tty_tolower)
+					/* switch to uppercase */
+					*op++ = _ebc_toupper[(int) *ip++];
+				else
+					/* switch to lowercase */
+					*op++ = _ebc_tolower[(int) *ip++];
+			else
+				/* no case switching, copy the character */
+				*op++ = *ip++;
+	}
+	/* return length of reformatted string. */
+	return op - buf;
+}
+
+static void sclp_get_input(struct gds_subvector *sv)
+{
+	unsigned char *str;
+	int count;
+
+	str = (unsigned char *) (sv + 1);
+	count = sv->length - sizeof(*sv);
+	if (sclp_tty_tolower)
+		EBC_TOLOWER(str, count);
+	count = sclp_switch_cases(str, count);
+	/* convert EBCDIC to ASCII (modify original input in SCCB) */
+	sclp_ebcasc_str(str, count);
+
+	/* transfer input to high level driver */
+	sclp_tty_input(str, count);
+}
+
+static inline void sclp_eval_selfdeftextmsg(struct gds_subvector *sv)
+{
+	void *end;
+
+	end = (void *) sv + sv->length;
+	for (sv = sv + 1; (void *) sv < end; sv = (void *) sv + sv->length)
+		if (sv->key == 0x30)
+			sclp_get_input(sv);
+}
+
+static inline void sclp_eval_textcmd(struct gds_vector *v)
+{
+	struct gds_subvector *sv;
+	void *end;
+
+	end = (void *) v + v->length;
+	for (sv = (struct gds_subvector *) (v + 1);
+	     (void *) sv < end; sv = (void *) sv + sv->length)
+		if (sv->key == GDS_KEY_SELFDEFTEXTMSG)
+			sclp_eval_selfdeftextmsg(sv);
+
+}
+
+static inline void sclp_eval_cpmsu(struct gds_vector *v)
+{
+	void *end;
+
+	end = (void *) v + v->length;
+	for (v = v + 1; (void *) v < end; v = (void *) v + v->length)
+		if (v->gds_id == GDS_ID_TEXTCMD)
+			sclp_eval_textcmd(v);
+}
+
+
+static inline void sclp_eval_mdsmu(struct gds_vector *v)
+{
+	v = sclp_find_gds_vector(v + 1, (void *) v + v->length, GDS_ID_CPMSU);
+	if (v)
+		sclp_eval_cpmsu(v);
+}
+
+static void sclp_tty_receiver(struct evbuf_header *evbuf)
+{
+	struct gds_vector *v;
+
+	v = sclp_find_gds_vector(evbuf + 1, (void *) evbuf + evbuf->length,
+				 GDS_ID_MDSMU);
+	if (v)
+		sclp_eval_mdsmu(v);
+}
+
+static void
+sclp_tty_state_change(struct sclp_register *reg)
+{
+}
+
+static struct sclp_register sclp_input_event =
+{
+	.receive_mask = EVTYP_OPCMD_MASK | EVTYP_PMSGCMD_MASK,
+	.state_change_fn = sclp_tty_state_change,
+	.receiver_fn = sclp_tty_receiver
+};
+
+static const struct tty_operations sclp_ops = {
+	.open = sclp_tty_open,
+	.close = sclp_tty_close,
+	.write = sclp_tty_write,
+	.put_char = sclp_tty_put_char,
+	.flush_chars = sclp_tty_flush_chars,
+	.write_room = sclp_tty_write_room,
+	.chars_in_buffer = sclp_tty_chars_in_buffer,
+	.flush_buffer = sclp_tty_flush_buffer,
+};
+
+static int __init
+sclp_tty_init(void)
+{
+	struct tty_driver *driver;
+	void *page;
+	int i;
+	int rc;
+
+	if (!CONSOLE_IS_SCLP)
+		return 0;
+	driver = alloc_tty_driver(1);
+	if (!driver)
+		return -ENOMEM;
+
+	rc = sclp_rw_init();
+	if (rc) {
+		put_tty_driver(driver);
+		return rc;
+	}
+	/* Allocate pages for output buffering */
+	INIT_LIST_HEAD(&sclp_tty_pages);
+	for (i = 0; i < MAX_KMEM_PAGES; i++) {
+		page = (void *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+		if (page == NULL) {
+			put_tty_driver(driver);
+			return -ENOMEM;
+		}
+		list_add_tail((struct list_head *) page, &sclp_tty_pages);
+	}
+	INIT_LIST_HEAD(&sclp_tty_outqueue);
+	spin_lock_init(&sclp_tty_lock);
+	init_timer(&sclp_tty_timer);
+	sclp_ttybuf = NULL;
+	sclp_tty_buffer_count = 0;
+	if (MACHINE_IS_VM) {
+		/*
+		 * save 4 characters for the CPU number
+		 * written at start of each line by VM/CP
+		 */
+		sclp_tty_columns = 76;
+		/* case input lines to lowercase */
+		sclp_tty_tolower = 1;
+	}
+	sclp_tty_chars_count = 0;
+
+	rc = sclp_register(&sclp_input_event);
+	if (rc) {
+		put_tty_driver(driver);
+		return rc;
+	}
+
+	tty_port_init(&sclp_port);
+
+	driver->driver_name = "sclp_line";
+	driver->name = "sclp_line";
+	driver->major = TTY_MAJOR;
+	driver->minor_start = 64;
+	driver->type = TTY_DRIVER_TYPE_SYSTEM;
+	driver->subtype = SYSTEM_TYPE_TTY;
+	driver->init_termios = tty_std_termios;
+	driver->init_termios.c_iflag = IGNBRK | IGNPAR;
+	driver->init_termios.c_oflag = ONLCR;
+	driver->init_termios.c_lflag = ISIG | ECHO;
+	driver->flags = TTY_DRIVER_REAL_RAW;
+	tty_set_operations(driver, &sclp_ops);
+	tty_port_link_device(&sclp_port, driver, 0);
+	rc = tty_register_driver(driver);
+	if (rc) {
+		put_tty_driver(driver);
+		tty_port_destroy(&sclp_port);
+		return rc;
+	}
+	sclp_tty_driver = driver;
+	return 0;
+}
+module_init(sclp_tty_init);

drivers/s390/char/sclp_tty.h

@@ -0,0 +1,17 @@
+/*
+ *    interface to the SCLP-read/write driver
+ *
+ *  S390 version
+ *    Copyright IBM Corp. 1999
+ *    Author(s): Martin Peschke <mpeschke@de.ibm.com>
+ *		 Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#ifndef __SCLP_TTY_H__
+#define __SCLP_TTY_H__
+
+#include <linux/tty_driver.h>
+
+extern struct tty_driver *sclp_tty_driver;
+
+#endif	/* __SCLP_TTY_H__ */

drivers/s390/char/sclp_vt220.c

@@ -0,0 +1,849 @@
+/*
+ * SCLP VT220 terminal driver.
+ *
+ * Copyright IBM Corp. 2003, 2009
+ *
+ * Author(s): Peter Oberparleiter <Peter.Oberparleiter@de.ibm.com>
+ */
+
+#include <linux/module.h>
+#include <linux/spinlock.h>
+#include <linux/list.h>
+#include <linux/wait.h>
+#include <linux/timer.h>
+#include <linux/kernel.h>
+#include <linux/tty.h>
+#include <linux/tty_driver.h>
+#include <linux/tty_flip.h>
+#include <linux/errno.h>
+#include <linux/mm.h>
+#include <linux/major.h>
+#include <linux/console.h>
+#include <linux/kdev_t.h>
+#include <linux/interrupt.h>
+#include <linux/init.h>
+#include <linux/reboot.h>
+#include <linux/slab.h>
+
+#include <asm/uaccess.h>
+#include "sclp.h"
+
+#define SCLP_VT220_MAJOR		TTY_MAJOR
+#define SCLP_VT220_MINOR		65
+#define SCLP_VT220_DRIVER_NAME		"sclp_vt220"
+#define SCLP_VT220_DEVICE_NAME		"ttysclp"
+#define SCLP_VT220_CONSOLE_NAME		"ttyS"
+#define SCLP_VT220_CONSOLE_INDEX	1	/* console=ttyS1 */
+
+/* Representation of a single write request */
+struct sclp_vt220_request {
+	struct list_head list;
+	struct sclp_req sclp_req;
+	int retry_count;
+};
+
+/* VT220 SCCB */
+struct sclp_vt220_sccb {
+	struct sccb_header header;
+	struct evbuf_header evbuf;
+};
+
+#define SCLP_VT220_MAX_CHARS_PER_BUFFER	(PAGE_SIZE - \
+					 sizeof(struct sclp_vt220_request) - \
+					 sizeof(struct sclp_vt220_sccb))
+
+/* Structures and data needed to register tty driver */
+static struct tty_driver *sclp_vt220_driver;
+
+static struct tty_port sclp_vt220_port;
+
+/* Lock to protect internal data from concurrent access */
+static spinlock_t sclp_vt220_lock;
+
+/* List of empty pages to be used as write request buffers */
+static struct list_head sclp_vt220_empty;
+
+/* List of pending requests */
+static struct list_head sclp_vt220_outqueue;
+
+/* Suspend mode flag */
+static int sclp_vt220_suspended;
+
+/* Flag that output queue is currently running */
+static int sclp_vt220_queue_running;
+
+/* Timer used for delaying write requests to merge subsequent messages into
+ * a single buffer */
+static struct timer_list sclp_vt220_timer;
+
+/* Pointer to current request buffer which has been partially filled but not
+ * yet sent */
+static struct sclp_vt220_request *sclp_vt220_current_request;
+
+/* Number of characters in current request buffer */
+static int sclp_vt220_buffered_chars;
+
+/* Counter controlling core driver initialization. */
+static int __initdata sclp_vt220_init_count;
+
+/* Flag indicating that sclp_vt220_current_request should really
+ * have been already queued but wasn't because the SCLP was processing
+ * another buffer */
+static int sclp_vt220_flush_later;
+
+static void sclp_vt220_receiver_fn(struct evbuf_header *evbuf);
+static void sclp_vt220_pm_event_fn(struct sclp_register *reg,
+				   enum sclp_pm_event sclp_pm_event);
+static int __sclp_vt220_emit(struct sclp_vt220_request *request);
+static void sclp_vt220_emit_current(void);
+
+/* Registration structure for SCLP output event buffers */
+static struct sclp_register sclp_vt220_register = {
+	.send_mask		= EVTYP_VT220MSG_MASK,
+	.pm_event_fn		= sclp_vt220_pm_event_fn,
+};
+
+/* Registration structure for SCLP input event buffers */
+static struct sclp_register sclp_vt220_register_input = {
+	.receive_mask		= EVTYP_VT220MSG_MASK,
+	.receiver_fn		= sclp_vt220_receiver_fn,
+};
+
+
+/*
+ * Put provided request buffer back into queue and check emit pending
+ * buffers if necessary.
+ */
+static void
+sclp_vt220_process_queue(struct sclp_vt220_request *request)
+{
+	unsigned long flags;
+	void *page;
+
+	do {
+		/* Put buffer back to list of empty buffers */
+		page = request->sclp_req.sccb;
+		spin_lock_irqsave(&sclp_vt220_lock, flags);
+		/* Move request from outqueue to empty queue */
+		list_del(&request->list);
+		list_add_tail((struct list_head *) page, &sclp_vt220_empty);
+		/* Check if there is a pending buffer on the out queue. */
+		request = NULL;
+		if (!list_empty(&sclp_vt220_outqueue))
+			request = list_entry(sclp_vt220_outqueue.next,
+					     struct sclp_vt220_request, list);
+		if (!request || sclp_vt220_suspended) {
+			sclp_vt220_queue_running = 0;
+			spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+			break;
+		}
+		spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+	} while (__sclp_vt220_emit(request));
+	if (request == NULL && sclp_vt220_flush_later)
+		sclp_vt220_emit_current();
+	tty_port_tty_wakeup(&sclp_vt220_port);
+}
+
+#define SCLP_BUFFER_MAX_RETRY		1
+
+/*
+ * Callback through which the result of a write request is reported by the
+ * SCLP.
+ */
+static void
+sclp_vt220_callback(struct sclp_req *request, void *data)
+{
+	struct sclp_vt220_request *vt220_request;
+	struct sclp_vt220_sccb *sccb;
+
+	vt220_request = (struct sclp_vt220_request *) data;
+	if (request->status == SCLP_REQ_FAILED) {
+		sclp_vt220_process_queue(vt220_request);
+		return;
+	}
+	sccb = (struct sclp_vt220_sccb *) vt220_request->sclp_req.sccb;
+
+	/* Check SCLP response code and choose suitable action	*/
+	switch (sccb->header.response_code) {
+	case 0x0020 :
+		break;
+
+	case 0x05f0: /* Target resource in improper state */
+		break;
+
+	case 0x0340: /* Contained SCLP equipment check */
+		if (++vt220_request->retry_count > SCLP_BUFFER_MAX_RETRY)
+			break;
+		/* Remove processed buffers and requeue rest */
+		if (sclp_remove_processed((struct sccb_header *) sccb) > 0) {
+			/* Not all buffers were processed */
+			sccb->header.response_code = 0x0000;
+			vt220_request->sclp_req.status = SCLP_REQ_FILLED;
+			if (sclp_add_request(request) == 0)
+				return;
+		}
+		break;
+
+	case 0x0040: /* SCLP equipment check */
+		if (++vt220_request->retry_count > SCLP_BUFFER_MAX_RETRY)
+			break;
+		sccb->header.response_code = 0x0000;
+		vt220_request->sclp_req.status = SCLP_REQ_FILLED;
+		if (sclp_add_request(request) == 0)
+			return;
+		break;
+
+	default:
+		break;
+	}
+	sclp_vt220_process_queue(vt220_request);
+}
+
+/*
+ * Emit vt220 request buffer to SCLP. Return zero on success, non-zero
+ * otherwise.
+ */
+static int
+__sclp_vt220_emit(struct sclp_vt220_request *request)
+{
+	request->sclp_req.command = SCLP_CMDW_WRITE_EVENT_DATA;
+	request->sclp_req.status = SCLP_REQ_FILLED;
+	request->sclp_req.callback = sclp_vt220_callback;
+	request->sclp_req.callback_data = (void *) request;
+
+	return sclp_add_request(&request->sclp_req);
+}
+
+/*
+ * Queue and emit current request.
+ */
+static void
+sclp_vt220_emit_current(void)
+{
+	unsigned long flags;
+	struct sclp_vt220_request *request;
+	struct sclp_vt220_sccb *sccb;
+
+	spin_lock_irqsave(&sclp_vt220_lock, flags);
+	if (sclp_vt220_current_request) {
+		sccb = (struct sclp_vt220_sccb *) 
+				sclp_vt220_current_request->sclp_req.sccb;
+		/* Only emit buffers with content */
+		if (sccb->header.length != sizeof(struct sclp_vt220_sccb)) {
+			list_add_tail(&sclp_vt220_current_request->list,
+				      &sclp_vt220_outqueue);
+			sclp_vt220_current_request = NULL;
+			if (timer_pending(&sclp_vt220_timer))
+				del_timer(&sclp_vt220_timer);
+		}
+		sclp_vt220_flush_later = 0;
+	}
+	if (sclp_vt220_queue_running || sclp_vt220_suspended)
+		goto out_unlock;
+	if (list_empty(&sclp_vt220_outqueue))
+		goto out_unlock;
+	request = list_first_entry(&sclp_vt220_outqueue,
+				   struct sclp_vt220_request, list);
+	sclp_vt220_queue_running = 1;
+	spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+
+	if (__sclp_vt220_emit(request))
+		sclp_vt220_process_queue(request);
+	return;
+out_unlock:
+	spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+}
+
+#define SCLP_NORMAL_WRITE	0x00
+
+/*
+ * Helper function to initialize a page with the sclp request structure.
+ */
+static struct sclp_vt220_request *
+sclp_vt220_initialize_page(void *page)
+{
+	struct sclp_vt220_request *request;
+	struct sclp_vt220_sccb *sccb;
+
+	/* Place request structure at end of page */
+	request = ((struct sclp_vt220_request *)
+			((addr_t) page + PAGE_SIZE)) - 1;
+	request->retry_count = 0;
+	request->sclp_req.sccb = page;
+	/* SCCB goes at start of page */
+	sccb = (struct sclp_vt220_sccb *) page;
+	memset((void *) sccb, 0, sizeof(struct sclp_vt220_sccb));
+	sccb->header.length = sizeof(struct sclp_vt220_sccb);
+	sccb->header.function_code = SCLP_NORMAL_WRITE;
+	sccb->header.response_code = 0x0000;
+	sccb->evbuf.type = EVTYP_VT220MSG;
+	sccb->evbuf.length = sizeof(struct evbuf_header);
+
+	return request;
+}
+
+static inline unsigned int
+sclp_vt220_space_left(struct sclp_vt220_request *request)
+{
+	struct sclp_vt220_sccb *sccb;
+	sccb = (struct sclp_vt220_sccb *) request->sclp_req.sccb;
+	return PAGE_SIZE - sizeof(struct sclp_vt220_request) -
+	       sccb->header.length;
+}
+
+static inline unsigned int
+sclp_vt220_chars_stored(struct sclp_vt220_request *request)
+{
+	struct sclp_vt220_sccb *sccb;
+	sccb = (struct sclp_vt220_sccb *) request->sclp_req.sccb;
+	return sccb->evbuf.length - sizeof(struct evbuf_header);
+}
+
+/*
+ * Add msg to buffer associated with request. Return the number of characters
+ * added.
+ */
+static int
+sclp_vt220_add_msg(struct sclp_vt220_request *request,
+		   const unsigned char *msg, int count, int convertlf)
+{
+	struct sclp_vt220_sccb *sccb;
+	void *buffer;
+	unsigned char c;
+	int from;
+	int to;
+
+	if (count > sclp_vt220_space_left(request))
+		count = sclp_vt220_space_left(request);
+	if (count <= 0)
+		return 0;
+
+	sccb = (struct sclp_vt220_sccb *) request->sclp_req.sccb;
+	buffer = (void *) ((addr_t) sccb + sccb->header.length);
+
+	if (convertlf) {
+		/* Perform Linefeed conversion (0x0a -> 0x0a 0x0d)*/
+		for (from=0, to=0;
+		     (from < count) && (to < sclp_vt220_space_left(request));
+		     from++) {
+			/* Retrieve character */
+			c = msg[from];
+			/* Perform conversion */
+			if (c == 0x0a) {
+				if (to + 1 < sclp_vt220_space_left(request)) {
+					((unsigned char *) buffer)[to++] = c;
+					((unsigned char *) buffer)[to++] = 0x0d;
+				} else
+					break;
+
+			} else
+				((unsigned char *) buffer)[to++] = c;
+		}
+		sccb->header.length += to;
+		sccb->evbuf.length += to;
+		return from;
+	} else {
+		memcpy(buffer, (const void *) msg, count);
+		sccb->header.length += count;
+		sccb->evbuf.length += count;
+		return count;
+	}
+}
+
+/*
+ * Emit buffer after having waited long enough for more data to arrive.
+ */
+static void
+sclp_vt220_timeout(unsigned long data)
+{
+	sclp_vt220_emit_current();
+}
+
+#define BUFFER_MAX_DELAY	HZ/20
+
+/*
+ * Drop oldest console buffer if sclp_con_drop is set
+ */
+static int
+sclp_vt220_drop_buffer(void)
+{
+	struct list_head *list;
+	struct sclp_vt220_request *request;
+	void *page;
+
+	if (!sclp_console_drop)
+		return 0;
+	list = sclp_vt220_outqueue.next;
+	if (sclp_vt220_queue_running)
+		/* The first element is in I/O */
+		list = list->next;
+	if (list == &sclp_vt220_outqueue)
+		return 0;
+	list_del(list);
+	request = list_entry(list, struct sclp_vt220_request, list);
+	page = request->sclp_req.sccb;
+	list_add_tail((struct list_head *) page, &sclp_vt220_empty);
+	return 1;
+}
+
+/* 
+ * Internal implementation of the write function. Write COUNT bytes of data
+ * from memory at BUF
+ * to the SCLP interface. In case that the data does not fit into the current
+ * write buffer, emit the current one and allocate a new one. If there are no
+ * more empty buffers available, wait until one gets emptied. If DO_SCHEDULE
+ * is non-zero, the buffer will be scheduled for emitting after a timeout -
+ * otherwise the user has to explicitly call the flush function.
+ * A non-zero CONVERTLF parameter indicates that 0x0a characters in the message
+ * buffer should be converted to 0x0a 0x0d. After completion, return the number
+ * of bytes written.
+ */
+static int
+__sclp_vt220_write(const unsigned char *buf, int count, int do_schedule,
+		   int convertlf, int may_fail)
+{
+	unsigned long flags;
+	void *page;
+	int written;
+	int overall_written;
+
+	if (count <= 0)
+		return 0;
+	overall_written = 0;
+	spin_lock_irqsave(&sclp_vt220_lock, flags);
+	do {
+		/* Create an sclp output buffer if none exists yet */
+		if (sclp_vt220_current_request == NULL) {
+			if (list_empty(&sclp_vt220_empty))
+				sclp_console_full++;
+			while (list_empty(&sclp_vt220_empty)) {
+				if (may_fail || sclp_vt220_suspended)
+					goto out;
+				if (sclp_vt220_drop_buffer())
+					break;
+				spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+
+				sclp_sync_wait();
+				spin_lock_irqsave(&sclp_vt220_lock, flags);
+			}
+			page = (void *) sclp_vt220_empty.next;
+			list_del((struct list_head *) page);
+			sclp_vt220_current_request =
+				sclp_vt220_initialize_page(page);
+		}
+		/* Try to write the string to the current request buffer */
+		written = sclp_vt220_add_msg(sclp_vt220_current_request,
+					     buf, count, convertlf);
+		overall_written += written;
+		if (written == count)
+			break;
+		/*
+		 * Not all characters could be written to the current
+		 * output buffer. Emit the buffer, create a new buffer
+		 * and then output the rest of the string.
+		 */
+		spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+		sclp_vt220_emit_current();
+		spin_lock_irqsave(&sclp_vt220_lock, flags);
+		buf += written;
+		count -= written;
+	} while (count > 0);
+	/* Setup timer to output current console buffer after some time */
+	if (sclp_vt220_current_request != NULL &&
+	    !timer_pending(&sclp_vt220_timer) && do_schedule) {
+		sclp_vt220_timer.function = sclp_vt220_timeout;
+		sclp_vt220_timer.data = 0UL;
+		sclp_vt220_timer.expires = jiffies + BUFFER_MAX_DELAY;
+		add_timer(&sclp_vt220_timer);
+	}
+out:
+	spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+	return overall_written;
+}
+
+/*
+ * This routine is called by the kernel to write a series of
+ * characters to the tty device.  The characters may come from
+ * user space or kernel space.  This routine will return the
+ * number of characters actually accepted for writing.
+ */
+static int
+sclp_vt220_write(struct tty_struct *tty, const unsigned char *buf, int count)
+{
+	return __sclp_vt220_write(buf, count, 1, 0, 1);
+}
+
+#define SCLP_VT220_SESSION_ENDED	0x01
+#define	SCLP_VT220_SESSION_STARTED	0x80
+#define SCLP_VT220_SESSION_DATA		0x00
+
+/*
+ * Called by the SCLP to report incoming event buffers.
+ */
+static void
+sclp_vt220_receiver_fn(struct evbuf_header *evbuf)
+{
+	char *buffer;
+	unsigned int count;
+
+	buffer = (char *) ((addr_t) evbuf + sizeof(struct evbuf_header));
+	count = evbuf->length - sizeof(struct evbuf_header);
+
+	switch (*buffer) {
+	case SCLP_VT220_SESSION_ENDED:
+	case SCLP_VT220_SESSION_STARTED:
+		break;
+	case SCLP_VT220_SESSION_DATA:
+		/* Send input to line discipline */
+		buffer++;
+		count--;
+		tty_insert_flip_string(&sclp_vt220_port, buffer, count);
+		tty_flip_buffer_push(&sclp_vt220_port);
+		break;
+	}
+}
+
+/*
+ * This routine is called when a particular tty device is opened.
+ */
+static int
+sclp_vt220_open(struct tty_struct *tty, struct file *filp)
+{
+	if (tty->count == 1) {
+		tty_port_tty_set(&sclp_vt220_port, tty);
+		sclp_vt220_port.low_latency = 0;
+		if (!tty->winsize.ws_row && !tty->winsize.ws_col) {
+			tty->winsize.ws_row = 24;
+			tty->winsize.ws_col = 80;
+		}
+	}
+	return 0;
+}
+
+/*
+ * This routine is called when a particular tty device is closed.
+ */
+static void
+sclp_vt220_close(struct tty_struct *tty, struct file *filp)
+{
+	if (tty->count == 1)
+		tty_port_tty_set(&sclp_vt220_port, NULL);
+}
+
+/*
+ * This routine is called by the kernel to write a single
+ * character to the tty device.  If the kernel uses this routine,
+ * it must call the flush_chars() routine (if defined) when it is
+ * done stuffing characters into the driver.
+ */
+static int
+sclp_vt220_put_char(struct tty_struct *tty, unsigned char ch)
+{
+	return __sclp_vt220_write(&ch, 1, 0, 0, 1);
+}
+
+/*
+ * This routine is called by the kernel after it has written a
+ * series of characters to the tty device using put_char().  
+ */
+static void
+sclp_vt220_flush_chars(struct tty_struct *tty)
+{
+	if (!sclp_vt220_queue_running)
+		sclp_vt220_emit_current();
+	else
+		sclp_vt220_flush_later = 1;
+}
+
+/*
+ * This routine returns the numbers of characters the tty driver
+ * will accept for queuing to be written.  This number is subject
+ * to change as output buffers get emptied, or if the output flow
+ * control is acted.
+ */
+static int
+sclp_vt220_write_room(struct tty_struct *tty)
+{
+	unsigned long flags;
+	struct list_head *l;
+	int count;
+
+	spin_lock_irqsave(&sclp_vt220_lock, flags);
+	count = 0;
+	if (sclp_vt220_current_request != NULL)
+		count = sclp_vt220_space_left(sclp_vt220_current_request);
+	list_for_each(l, &sclp_vt220_empty)
+		count += SCLP_VT220_MAX_CHARS_PER_BUFFER;
+	spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+	return count;
+}
+
+/*
+ * Return number of buffered chars.
+ */
+static int
+sclp_vt220_chars_in_buffer(struct tty_struct *tty)
+{
+	unsigned long flags;
+	struct list_head *l;
+	struct sclp_vt220_request *r;
+	int count;
+
+	spin_lock_irqsave(&sclp_vt220_lock, flags);
+	count = 0;
+	if (sclp_vt220_current_request != NULL)
+		count = sclp_vt220_chars_stored(sclp_vt220_current_request);
+	list_for_each(l, &sclp_vt220_outqueue) {
+		r = list_entry(l, struct sclp_vt220_request, list);
+		count += sclp_vt220_chars_stored(r);
+	}
+	spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+	return count;
+}
+
+/*
+ * Pass on all buffers to the hardware. Return only when there are no more
+ * buffers pending.
+ */
+static void
+sclp_vt220_flush_buffer(struct tty_struct *tty)
+{
+	sclp_vt220_emit_current();
+}
+
+/* Release allocated pages. */
+static void __init __sclp_vt220_free_pages(void)
+{
+	struct list_head *page, *p;
+
+	list_for_each_safe(page, p, &sclp_vt220_empty) {
+		list_del(page);
+		free_page((unsigned long) page);
+	}
+}
+
+/* Release memory and unregister from sclp core. Controlled by init counting -
+ * only the last invoker will actually perform these actions. */
+static void __init __sclp_vt220_cleanup(void)
+{
+	sclp_vt220_init_count--;
+	if (sclp_vt220_init_count != 0)
+		return;
+	sclp_unregister(&sclp_vt220_register);
+	__sclp_vt220_free_pages();
+	tty_port_destroy(&sclp_vt220_port);
+}
+
+/* Allocate buffer pages and register with sclp core. Controlled by init
+ * counting - only the first invoker will actually perform these actions. */
+static int __init __sclp_vt220_init(int num_pages)
+{
+	void *page;
+	int i;
+	int rc;
+
+	sclp_vt220_init_count++;
+	if (sclp_vt220_init_count != 1)
+		return 0;
+	spin_lock_init(&sclp_vt220_lock);
+	INIT_LIST_HEAD(&sclp_vt220_empty);
+	INIT_LIST_HEAD(&sclp_vt220_outqueue);
+	init_timer(&sclp_vt220_timer);
+	tty_port_init(&sclp_vt220_port);
+	sclp_vt220_current_request = NULL;
+	sclp_vt220_buffered_chars = 0;
+	sclp_vt220_flush_later = 0;
+
+	/* Allocate pages for output buffering */
+	rc = -ENOMEM;
+	for (i = 0; i < num_pages; i++) {
+		page = (void *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
+		if (!page)
+			goto out;
+		list_add_tail(page, &sclp_vt220_empty);
+	}
+	rc = sclp_register(&sclp_vt220_register);
+out:
+	if (rc) {
+		__sclp_vt220_free_pages();
+		sclp_vt220_init_count--;
+		tty_port_destroy(&sclp_vt220_port);
+	}
+	return rc;
+}
+
+static const struct tty_operations sclp_vt220_ops = {
+	.open = sclp_vt220_open,
+	.close = sclp_vt220_close,
+	.write = sclp_vt220_write,
+	.put_char = sclp_vt220_put_char,
+	.flush_chars = sclp_vt220_flush_chars,
+	.write_room = sclp_vt220_write_room,
+	.chars_in_buffer = sclp_vt220_chars_in_buffer,
+	.flush_buffer = sclp_vt220_flush_buffer,
+};
+
+/*
+ * Register driver with SCLP and Linux and initialize internal tty structures.
+ */
+static int __init sclp_vt220_tty_init(void)
+{
+	struct tty_driver *driver;
+	int rc;
+
+	/* Note: we're not testing for CONSOLE_IS_SCLP here to preserve
+	 * symmetry between VM and LPAR systems regarding ttyS1. */
+	driver = alloc_tty_driver(1);
+	if (!driver)
+		return -ENOMEM;
+	rc = __sclp_vt220_init(MAX_KMEM_PAGES);
+	if (rc)
+		goto out_driver;
+
+	driver->driver_name = SCLP_VT220_DRIVER_NAME;
+	driver->name = SCLP_VT220_DEVICE_NAME;
+	driver->major = SCLP_VT220_MAJOR;
+	driver->minor_start = SCLP_VT220_MINOR;
+	driver->type = TTY_DRIVER_TYPE_SYSTEM;
+	driver->subtype = SYSTEM_TYPE_TTY;
+	driver->init_termios = tty_std_termios;
+	driver->flags = TTY_DRIVER_REAL_RAW;
+	tty_set_operations(driver, &sclp_vt220_ops);
+	tty_port_link_device(&sclp_vt220_port, driver, 0);
+
+	rc = tty_register_driver(driver);
+	if (rc)
+		goto out_init;
+	rc = sclp_register(&sclp_vt220_register_input);
+	if (rc)
+		goto out_reg;
+	sclp_vt220_driver = driver;
+	return 0;
+
+out_reg:
+	tty_unregister_driver(driver);
+out_init:
+	__sclp_vt220_cleanup();
+out_driver:
+	put_tty_driver(driver);
+	return rc;
+}
+__initcall(sclp_vt220_tty_init);
+
+static void __sclp_vt220_flush_buffer(void)
+{
+	unsigned long flags;
+
+	sclp_vt220_emit_current();
+	spin_lock_irqsave(&sclp_vt220_lock, flags);
+	if (timer_pending(&sclp_vt220_timer))
+		del_timer(&sclp_vt220_timer);
+	while (sclp_vt220_queue_running) {
+		spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+		sclp_sync_wait();
+		spin_lock_irqsave(&sclp_vt220_lock, flags);
+	}
+	spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+}
+
+/*
+ * Resume console: If there are cached messages, emit them.
+ */
+static void sclp_vt220_resume(void)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&sclp_vt220_lock, flags);
+	sclp_vt220_suspended = 0;
+	spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+	sclp_vt220_emit_current();
+}
+
+/*
+ * Suspend console: Set suspend flag and flush console
+ */
+static void sclp_vt220_suspend(void)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&sclp_vt220_lock, flags);
+	sclp_vt220_suspended = 1;
+	spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+	__sclp_vt220_flush_buffer();
+}
+
+static void sclp_vt220_pm_event_fn(struct sclp_register *reg,
+				   enum sclp_pm_event sclp_pm_event)
+{
+	switch (sclp_pm_event) {
+	case SCLP_PM_EVENT_FREEZE:
+		sclp_vt220_suspend();
+		break;
+	case SCLP_PM_EVENT_RESTORE:
+	case SCLP_PM_EVENT_THAW:
+		sclp_vt220_resume();
+		break;
+	}
+}
+
+#ifdef CONFIG_SCLP_VT220_CONSOLE
+
+static void
+sclp_vt220_con_write(struct console *con, const char *buf, unsigned int count)
+{
+	__sclp_vt220_write((const unsigned char *) buf, count, 1, 1, 0);
+}
+
+static struct tty_driver *
+sclp_vt220_con_device(struct console *c, int *index)
+{
+	*index = 0;
+	return sclp_vt220_driver;
+}
+
+static int
+sclp_vt220_notify(struct notifier_block *self,
+			  unsigned long event, void *data)
+{
+	__sclp_vt220_flush_buffer();
+	return NOTIFY_OK;
+}
+
+static struct notifier_block on_panic_nb = {
+	.notifier_call = sclp_vt220_notify,
+	.priority = 1,
+};
+
+static struct notifier_block on_reboot_nb = {
+	.notifier_call = sclp_vt220_notify,
+	.priority = 1,
+};
+
+/* Structure needed to register with printk */
+static struct console sclp_vt220_console =
+{
+	.name = SCLP_VT220_CONSOLE_NAME,
+	.write = sclp_vt220_con_write,
+	.device = sclp_vt220_con_device,
+	.flags = CON_PRINTBUFFER,
+	.index = SCLP_VT220_CONSOLE_INDEX
+};
+
+static int __init
+sclp_vt220_con_init(void)
+{
+	int rc;
+
+	rc = __sclp_vt220_init(sclp_console_pages);
+	if (rc)
+		return rc;
+	/* Attach linux console */
+	atomic_notifier_chain_register(&panic_notifier_list, &on_panic_nb);
+	register_reboot_notifier(&on_reboot_nb);
+	register_console(&sclp_vt220_console);
+	return 0;
+}
+
+console_initcall(sclp_vt220_con_init);
+#endif /* CONFIG_SCLP_VT220_CONSOLE */
+

drivers/s390/char/tape.h

@@ -0,0 +1,369 @@
+/*
+ *    tape device driver for 3480/3490E/3590 tapes.
+ *
+ *  S390 and zSeries version
+ *    Copyright IBM Corp. 2001, 2009
+ *    Author(s): Carsten Otte <cotte@de.ibm.com>
+ *		 Tuan Ngo-Anh <ngoanh@de.ibm.com>
+ *		 Martin Schwidefsky <schwidefsky@de.ibm.com>
+ *		 Stefan Bader <shbader@de.ibm.com>
+ */
+
+#ifndef _TAPE_H
+#define _TAPE_H
+
+#include <asm/ccwdev.h>
+#include <asm/debug.h>
+#include <asm/idals.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/mtio.h>
+#include <linux/interrupt.h>
+#include <linux/workqueue.h>
+
+struct gendisk;
+
+/*
+ * Define DBF_LIKE_HELL for lots of messages in the debug feature.
+ */
+#define DBF_LIKE_HELL
+#ifdef  DBF_LIKE_HELL
+#define DBF_LH(level, str, ...) \
+do { \
+	debug_sprintf_event(TAPE_DBF_AREA, level, str, ## __VA_ARGS__); \
+} while (0)
+#else
+#define DBF_LH(level, str, ...) do {} while(0)
+#endif
+
+/*
+ * macros s390 debug feature (dbf)
+ */
+#define DBF_EVENT(d_level, d_str...) \
+do { \
+	debug_sprintf_event(TAPE_DBF_AREA, d_level, d_str); \
+} while (0)
+
+#define DBF_EXCEPTION(d_level, d_str...) \
+do { \
+	debug_sprintf_exception(TAPE_DBF_AREA, d_level, d_str); \
+} while (0)
+
+#define TAPE_VERSION_MAJOR 2
+#define TAPE_VERSION_MINOR 0
+#define TAPE_MAGIC "tape"
+
+#define TAPE_MINORS_PER_DEV 2	    /* two minors per device */
+#define TAPEBLOCK_HSEC_SIZE	2048
+#define TAPEBLOCK_HSEC_S2B	2
+#define TAPEBLOCK_RETRIES	5
+
+enum tape_medium_state {
+	MS_UNKNOWN,
+	MS_LOADED,
+	MS_UNLOADED,
+	MS_SIZE
+};
+
+enum tape_state {
+	TS_UNUSED=0,
+	TS_IN_USE,
+	TS_BLKUSE,
+	TS_INIT,
+	TS_NOT_OPER,
+	TS_SIZE
+};
+
+enum tape_op {
+	TO_BLOCK,	/* Block read */
+	TO_BSB,		/* Backward space block */
+	TO_BSF,		/* Backward space filemark */
+	TO_DSE,		/* Data security erase */
+	TO_FSB,		/* Forward space block */
+	TO_FSF,		/* Forward space filemark */
+	TO_LBL,		/* Locate block label */
+	TO_NOP,		/* No operation */
+	TO_RBA,		/* Read backward */
+	TO_RBI,		/* Read block information */
+	TO_RFO,		/* Read forward */
+	TO_REW,		/* Rewind tape */
+	TO_RUN,		/* Rewind and unload tape */
+	TO_WRI,		/* Write block */
+	TO_WTM,		/* Write tape mark */
+	TO_MSEN,	/* Medium sense */
+	TO_LOAD,	/* Load tape */
+	TO_READ_CONFIG, /* Read configuration data */
+	TO_READ_ATTMSG, /* Read attention message */
+	TO_DIS,		/* Tape display */
+	TO_ASSIGN,	/* Assign tape to channel path */
+	TO_UNASSIGN,	/* Unassign tape from channel path */
+	TO_CRYPT_ON,	/* Enable encrpytion */
+	TO_CRYPT_OFF,	/* Disable encrpytion */
+	TO_KEKL_SET,	/* Set KEK label */
+	TO_KEKL_QUERY,	/* Query KEK label */
+	TO_RDC,		/* Read device characteristics */
+	TO_SIZE,	/* #entries in tape_op_t */
+};
+
+/* Forward declaration */
+struct tape_device;
+
+/* tape_request->status can be: */
+enum tape_request_status {
+	TAPE_REQUEST_INIT,	/* request is ready to be processed */
+	TAPE_REQUEST_QUEUED,	/* request is queued to be processed */
+	TAPE_REQUEST_IN_IO,	/* request is currently in IO */
+	TAPE_REQUEST_DONE,	/* request is completed. */
+	TAPE_REQUEST_CANCEL,	/* request should be canceled. */
+	TAPE_REQUEST_LONG_BUSY, /* request has to be restarted after long busy */
+};
+
+/* Tape CCW request */
+struct tape_request {
+	struct list_head list;		/* list head for request queueing. */
+	struct tape_device *device;	/* tape device of this request */
+	struct ccw1 *cpaddr;		/* address of the channel program. */
+	void *cpdata;			/* pointer to ccw data. */
+	enum tape_request_status status;/* status of this request */
+	int options;			/* options for execution. */
+	int retries;			/* retry counter for error recovery. */
+	int rescnt;			/* residual count from devstat. */
+
+	/* Callback for delivering final status. */
+	void (*callback)(struct tape_request *, void *);
+	void *callback_data;
+
+	enum tape_op op;
+	int rc;
+};
+
+/* Function type for magnetic tape commands */
+typedef int (*tape_mtop_fn)(struct tape_device *, int);
+
+/* Size of the array containing the mtops for a discipline */
+#define TAPE_NR_MTOPS (MTMKPART+1)
+
+/* Tape Discipline */
+struct tape_discipline {
+	struct module *owner;
+	int  (*setup_device)(struct tape_device *);
+	void (*cleanup_device)(struct tape_device *);
+	int (*irq)(struct tape_device *, struct tape_request *, struct irb *);
+	struct tape_request *(*read_block)(struct tape_device *, size_t);
+	struct tape_request *(*write_block)(struct tape_device *, size_t);
+	void (*process_eov)(struct tape_device*);
+	/* ioctl function for additional ioctls. */
+	int (*ioctl_fn)(struct tape_device *, unsigned int, unsigned long);
+	/* Array of tape commands with TAPE_NR_MTOPS entries */
+	tape_mtop_fn *mtop_array;
+};
+
+/*
+ * The discipline irq function either returns an error code (<0) which
+ * means that the request has failed with an error or one of the following:
+ */
+#define TAPE_IO_SUCCESS		0	/* request successful */
+#define TAPE_IO_PENDING		1	/* request still running */
+#define TAPE_IO_RETRY		2	/* retry to current request */
+#define TAPE_IO_STOP		3	/* stop the running request */
+#define TAPE_IO_LONG_BUSY	4	/* delay the running request */
+
+/* Char Frontend Data */
+struct tape_char_data {
+	struct idal_buffer *idal_buf;	/* idal buffer for user char data */
+	int block_size;			/*   of size block_size. */
+};
+
+/* Tape Info */
+struct tape_device {
+	/* entry in tape_device_list */
+	struct list_head		node;
+
+	int				cdev_id;
+	struct ccw_device *		cdev;
+	struct tape_class_device *	nt;
+	struct tape_class_device *	rt;
+
+	/* Device mutex to serialize tape commands. */
+	struct mutex			mutex;
+
+	/* Device discipline information. */
+	struct tape_discipline *	discipline;
+	void *				discdata;
+
+	/* Generic status flags */
+	long				tape_generic_status;
+
+	/* Device state information. */
+	wait_queue_head_t		state_change_wq;
+	enum tape_state			tape_state;
+	enum tape_medium_state		medium_state;
+	unsigned char *			modeset_byte;
+
+	/* Reference count. */
+	atomic_t			ref_count;
+
+	/* Request queue. */
+	struct list_head		req_queue;
+
+	/* Request wait queue. */
+	wait_queue_head_t		wait_queue;
+
+	/* Each tape device has (currently) two minor numbers. */
+	int				first_minor;
+
+	/* Number of tapemarks required for correct termination. */
+	int				required_tapemarks;
+
+	/* Block ID of the BOF */
+	unsigned int			bof;
+
+	/* Character device frontend data */
+	struct tape_char_data		char_data;
+
+	/* Function to start or stop the next request later. */
+	struct delayed_work		tape_dnr;
+
+	/* Timer for long busy */
+	struct timer_list		lb_timeout;
+
+};
+
+/* Externals from tape_core.c */
+extern struct tape_request *tape_alloc_request(int cplength, int datasize);
+extern void tape_free_request(struct tape_request *);
+extern int tape_do_io(struct tape_device *, struct tape_request *);
+extern int tape_do_io_async(struct tape_device *, struct tape_request *);
+extern int tape_do_io_interruptible(struct tape_device *, struct tape_request *);
+extern int tape_cancel_io(struct tape_device *, struct tape_request *);
+void tape_hotplug_event(struct tape_device *, int major, int action);
+
+static inline int
+tape_do_io_free(struct tape_device *device, struct tape_request *request)
+{
+	int rc;
+
+	rc = tape_do_io(device, request);
+	tape_free_request(request);
+	return rc;
+}
+
+static inline void
+tape_do_io_async_free(struct tape_device *device, struct tape_request *request)
+{
+	request->callback = (void *) tape_free_request;
+	request->callback_data = NULL;
+	tape_do_io_async(device, request);
+}
+
+extern int tape_oper_handler(int irq, int status);
+extern void tape_noper_handler(int irq, int status);
+extern int tape_open(struct tape_device *);
+extern int tape_release(struct tape_device *);
+extern int tape_mtop(struct tape_device *, int, int);
+extern void tape_state_set(struct tape_device *, enum tape_state);
+
+extern int tape_generic_online(struct tape_device *, struct tape_discipline *);
+extern int tape_generic_offline(struct ccw_device *);
+extern int tape_generic_pm_suspend(struct ccw_device *);
+
+/* Externals from tape_devmap.c */
+extern int tape_generic_probe(struct ccw_device *);
+extern void tape_generic_remove(struct ccw_device *);
+
+extern struct tape_device *tape_find_device(int devindex);
+extern struct tape_device *tape_get_device(struct tape_device *);
+extern void tape_put_device(struct tape_device *);
+
+/* Externals from tape_char.c */
+extern int tapechar_init(void);
+extern void tapechar_exit(void);
+extern int  tapechar_setup_device(struct tape_device *);
+extern void tapechar_cleanup_device(struct tape_device *);
+
+/* tape initialisation functions */
+#ifdef CONFIG_PROC_FS
+extern void tape_proc_init (void);
+extern void tape_proc_cleanup (void);
+#else
+static inline void tape_proc_init (void) {;}
+static inline void tape_proc_cleanup (void) {;}
+#endif
+
+/* a function for dumping device sense info */
+extern void tape_dump_sense_dbf(struct tape_device *, struct tape_request *,
+				struct irb *);
+
+/* functions for handling the status of a device */
+extern void tape_med_state_set(struct tape_device *, enum tape_medium_state);
+
+/* The debug area */
+extern debug_info_t *TAPE_DBF_AREA;
+
+/* functions for building ccws */
+static inline struct ccw1 *
+tape_ccw_cc(struct ccw1 *ccw, __u8 cmd_code, __u16 memsize, void *cda)
+{
+	ccw->cmd_code = cmd_code;
+	ccw->flags = CCW_FLAG_CC;
+	ccw->count = memsize;
+	ccw->cda = (__u32)(addr_t) cda;
+	return ccw + 1;
+}
+
+static inline struct ccw1 *
+tape_ccw_end(struct ccw1 *ccw, __u8 cmd_code, __u16 memsize, void *cda)
+{
+	ccw->cmd_code = cmd_code;
+	ccw->flags = 0;
+	ccw->count = memsize;
+	ccw->cda = (__u32)(addr_t) cda;
+	return ccw + 1;
+}
+
+static inline struct ccw1 *
+tape_ccw_cmd(struct ccw1 *ccw, __u8 cmd_code)
+{
+	ccw->cmd_code = cmd_code;
+	ccw->flags = 0;
+	ccw->count = 0;
+	ccw->cda = (__u32)(addr_t) &ccw->cmd_code;
+	return ccw + 1;
+}
+
+static inline struct ccw1 *
+tape_ccw_repeat(struct ccw1 *ccw, __u8 cmd_code, int count)
+{
+	while (count-- > 0) {
+		ccw->cmd_code = cmd_code;
+		ccw->flags = CCW_FLAG_CC;
+		ccw->count = 0;
+		ccw->cda = (__u32)(addr_t) &ccw->cmd_code;
+		ccw++;
+	}
+	return ccw;
+}
+
+static inline struct ccw1 *
+tape_ccw_cc_idal(struct ccw1 *ccw, __u8 cmd_code, struct idal_buffer *idal)
+{
+	ccw->cmd_code = cmd_code;
+	ccw->flags    = CCW_FLAG_CC;
+	idal_buffer_set_cda(idal, ccw);
+	return ccw++;
+}
+
+static inline struct ccw1 *
+tape_ccw_end_idal(struct ccw1 *ccw, __u8 cmd_code, struct idal_buffer *idal)
+{
+	ccw->cmd_code = cmd_code;
+	ccw->flags    = 0;
+	idal_buffer_set_cda(idal, ccw);
+	return ccw++;
+}
+
+/* Global vars */
+extern const char *tape_state_verbose[];
+extern const char *tape_op_verbose[];
+
+#endif /* for ifdef tape.h */

drivers/s390/char/tape_3590.h

@@ -0,0 +1,174 @@
+/*
+ *    tape device discipline for 3590 tapes.
+ *
+ *    Copyright IBM Corp. 2001, 2006
+ *    Author(s): Stefan Bader <shbader@de.ibm.com>
+ *		 Michael Holzheu <holzheu@de.ibm.com>
+ *		 Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#ifndef _TAPE_3590_H
+#define _TAPE_3590_H
+
+#define MEDIUM_SENSE	0xc2
+#define READ_PREVIOUS	0x0a
+#define MODE_SENSE	0xcf
+#define PERFORM_SS_FUNC 0x77
+#define READ_SS_DATA	0x3e
+
+#define PREP_RD_SS_DATA 0x18
+#define RD_ATTMSG	0x3
+
+#define SENSE_BRA_PER  0
+#define SENSE_BRA_CONT 1
+#define SENSE_BRA_RE   2
+#define SENSE_BRA_DRE  3
+
+#define SENSE_FMT_LIBRARY	0x23
+#define SENSE_FMT_UNSOLICITED	0x40
+#define SENSE_FMT_COMMAND_REJ	0x41
+#define SENSE_FMT_COMMAND_EXEC0 0x50
+#define SENSE_FMT_COMMAND_EXEC1 0x51
+#define SENSE_FMT_EVENT0	0x60
+#define SENSE_FMT_EVENT1	0x61
+#define SENSE_FMT_MIM		0x70
+#define SENSE_FMT_SIM		0x71
+
+#define MSENSE_UNASSOCIATED	 0x00
+#define MSENSE_ASSOCIATED_MOUNT	 0x01
+#define MSENSE_ASSOCIATED_UMOUNT 0x02
+#define MSENSE_CRYPT_MASK	 0x00000010
+
+#define TAPE_3590_MAX_MSG	 0xb0
+
+/* Datatypes */
+
+struct tape_3590_disc_data {
+	struct tape390_crypt_info crypt_info;
+	int read_back_op;
+};
+
+#define TAPE_3590_CRYPT_INFO(device) \
+	((struct tape_3590_disc_data*)(device->discdata))->crypt_info
+#define TAPE_3590_READ_BACK_OP(device) \
+	((struct tape_3590_disc_data*)(device->discdata))->read_back_op
+
+struct tape_3590_sense {
+
+	unsigned int command_rej:1;
+	unsigned int interv_req:1;
+	unsigned int bus_out_check:1;
+	unsigned int eq_check:1;
+	unsigned int data_check:1;
+	unsigned int overrun:1;
+	unsigned int def_unit_check:1;
+	unsigned int assgnd_elsew:1;
+
+	unsigned int locate_fail:1;
+	unsigned int inst_online:1;
+	unsigned int reserved:1;
+	unsigned int blk_seq_err:1;
+	unsigned int begin_part:1;
+	unsigned int wr_mode:1;
+	unsigned int wr_prot:1;
+	unsigned int not_cap:1;
+
+	unsigned int bra:2;
+	unsigned int lc:3;
+	unsigned int vlf_active:1;
+	unsigned int stm:1;
+	unsigned int med_pos:1;
+
+	unsigned int rac:8;
+
+	unsigned int rc_rqc:16;
+
+	unsigned int mc:8;
+
+	unsigned int sense_fmt:8;
+
+	union {
+		struct {
+			unsigned int emc:4;
+			unsigned int smc:4;
+			unsigned int sev:2;
+			unsigned int reserved:6;
+			unsigned int md:8;
+			unsigned int refcode:8;
+			unsigned int mid:16;
+			unsigned int mp:16;
+			unsigned char volid[6];
+			unsigned int fid:8;
+		} f70;
+		struct {
+			unsigned int emc:4;
+			unsigned int smc:4;
+			unsigned int sev:2;
+			unsigned int reserved1:5;
+			unsigned int mdf:1;
+			unsigned char md[3];
+			unsigned int simid:8;
+			unsigned int uid:16;
+			unsigned int refcode1:16;
+			unsigned int refcode2:16;
+			unsigned int refcode3:16;
+			unsigned int reserved2:8;
+		} f71;
+		unsigned char data[14];
+	} fmt;
+	unsigned char pad[10];
+
+} __attribute__ ((packed));
+
+struct tape_3590_med_sense {
+	unsigned int macst:4;
+	unsigned int masst:4;
+	char pad1[7];
+	unsigned int flags;
+	char pad2[116];
+} __attribute__ ((packed));
+
+struct tape_3590_rdc_data {
+	char data[64];
+} __attribute__ ((packed));
+
+/* Datastructures for 3592 encryption support */
+
+struct tape3592_kekl {
+	__u8 flags;
+	char label[64];
+} __attribute__ ((packed));
+
+struct tape3592_kekl_pair {
+	__u8 count;
+	struct tape3592_kekl kekl[2];
+} __attribute__ ((packed));
+
+struct tape3592_kekl_query_data {
+	__u16 len;
+	__u8  fmt;
+	__u8  mc;
+	__u32 id;
+	__u8  flags;
+	struct tape3592_kekl_pair kekls;
+	char reserved[116];
+} __attribute__ ((packed));
+
+struct tape3592_kekl_query_order {
+	__u8 code;
+	__u8 flags;
+	char reserved1[2];
+	__u8 max_count;
+	char reserved2[35];
+} __attribute__ ((packed));
+
+struct tape3592_kekl_set_order {
+	__u8 code;
+	__u8 flags;
+	char reserved1[2];
+	__u8 op;
+	struct tape3592_kekl_pair kekls;
+	char reserved2[120];
+} __attribute__ ((packed));
+
+#endif /* _TAPE_3590_H */

drivers/s390/char/tape_char.c

@@ -0,0 +1,500 @@
+/*
+ *    character device frontend for tape device driver
+ *
+ *  S390 and zSeries version
+ *    Copyright IBM Corp. 2001, 2006
+ *    Author(s): Carsten Otte <cotte@de.ibm.com>
+ *		 Michael Holzheu <holzheu@de.ibm.com>
+ *		 Tuan Ngo-Anh <ngoanh@de.ibm.com>
+ *		 Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#define KMSG_COMPONENT "tape"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/proc_fs.h>
+#include <linux/mtio.h>
+#include <linux/compat.h>
+
+#include <asm/uaccess.h>
+
+#define TAPE_DBF_AREA	tape_core_dbf
+
+#include "tape.h"
+#include "tape_std.h"
+#include "tape_class.h"
+
+#define TAPECHAR_MAJOR		0	/* get dynamic major */
+
+/*
+ * file operation structure for tape character frontend
+ */
+static ssize_t tapechar_read(struct file *, char __user *, size_t, loff_t *);
+static ssize_t tapechar_write(struct file *, const char __user *, size_t, loff_t *);
+static int tapechar_open(struct inode *,struct file *);
+static int tapechar_release(struct inode *,struct file *);
+static long tapechar_ioctl(struct file *, unsigned int, unsigned long);
+#ifdef CONFIG_COMPAT
+static long tapechar_compat_ioctl(struct file *, unsigned int, unsigned long);
+#endif
+
+static const struct file_operations tape_fops =
+{
+	.owner = THIS_MODULE,
+	.read = tapechar_read,
+	.write = tapechar_write,
+	.unlocked_ioctl = tapechar_ioctl,
+#ifdef CONFIG_COMPAT
+	.compat_ioctl = tapechar_compat_ioctl,
+#endif
+	.open = tapechar_open,
+	.release = tapechar_release,
+	.llseek = no_llseek,
+};
+
+static int tapechar_major = TAPECHAR_MAJOR;
+
+/*
+ * This function is called for every new tapedevice
+ */
+int
+tapechar_setup_device(struct tape_device * device)
+{
+	char	device_name[20];
+
+	sprintf(device_name, "ntibm%i", device->first_minor / 2);
+	device->nt = register_tape_dev(
+		&device->cdev->dev,
+		MKDEV(tapechar_major, device->first_minor),
+		&tape_fops,
+		device_name,
+		"non-rewinding"
+	);
+	device_name[0] = 'r';
+	device->rt = register_tape_dev(
+		&device->cdev->dev,
+		MKDEV(tapechar_major, device->first_minor + 1),
+		&tape_fops,
+		device_name,
+		"rewinding"
+	);
+
+	return 0;
+}
+
+void
+tapechar_cleanup_device(struct tape_device *device)
+{
+	unregister_tape_dev(&device->cdev->dev, device->rt);
+	device->rt = NULL;
+	unregister_tape_dev(&device->cdev->dev, device->nt);
+	device->nt = NULL;
+}
+
+static int
+tapechar_check_idalbuffer(struct tape_device *device, size_t block_size)
+{
+	struct idal_buffer *new;
+
+	if (device->char_data.idal_buf != NULL &&
+	    device->char_data.idal_buf->size == block_size)
+		return 0;
+
+	if (block_size > MAX_BLOCKSIZE) {
+		DBF_EVENT(3, "Invalid blocksize (%zd > %d)\n",
+			block_size, MAX_BLOCKSIZE);
+		return -EINVAL;
+	}
+
+	/* The current idal buffer is not correct. Allocate a new one. */
+	new = idal_buffer_alloc(block_size, 0);
+	if (IS_ERR(new))
+		return -ENOMEM;
+
+	if (device->char_data.idal_buf != NULL)
+		idal_buffer_free(device->char_data.idal_buf);
+
+	device->char_data.idal_buf = new;
+
+	return 0;
+}
+
+/*
+ * Tape device read function
+ */
+static ssize_t
+tapechar_read(struct file *filp, char __user *data, size_t count, loff_t *ppos)
+{
+	struct tape_device *device;
+	struct tape_request *request;
+	size_t block_size;
+	int rc;
+
+	DBF_EVENT(6, "TCHAR:read\n");
+	device = (struct tape_device *) filp->private_data;
+
+	/*
+	 * If the tape isn't terminated yet, do it now. And since we then
+	 * are at the end of the tape there wouldn't be anything to read
+	 * anyways. So we return immediately.
+	 */
+	if(device->required_tapemarks) {
+		return tape_std_terminate_write(device);
+	}
+
+	/* Find out block size to use */
+	if (device->char_data.block_size != 0) {
+		if (count < device->char_data.block_size) {
+			DBF_EVENT(3, "TCHAR:read smaller than block "
+				  "size was requested\n");
+			return -EINVAL;
+		}
+		block_size = device->char_data.block_size;
+	} else {
+		block_size = count;
+	}
+
+	rc = tapechar_check_idalbuffer(device, block_size);
+	if (rc)
+		return rc;
+
+	DBF_EVENT(6, "TCHAR:nbytes: %lx\n", block_size);
+	/* Let the discipline build the ccw chain. */
+	request = device->discipline->read_block(device, block_size);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	/* Execute it. */
+	rc = tape_do_io(device, request);
+	if (rc == 0) {
+		rc = block_size - request->rescnt;
+		DBF_EVENT(6, "TCHAR:rbytes:  %x\n", rc);
+		/* Copy data from idal buffer to user space. */
+		if (idal_buffer_to_user(device->char_data.idal_buf,
+					data, rc) != 0)
+			rc = -EFAULT;
+	}
+	tape_free_request(request);
+	return rc;
+}
+
+/*
+ * Tape device write function
+ */
+static ssize_t
+tapechar_write(struct file *filp, const char __user *data, size_t count, loff_t *ppos)
+{
+	struct tape_device *device;
+	struct tape_request *request;
+	size_t block_size;
+	size_t written;
+	int nblocks;
+	int i, rc;
+
+	DBF_EVENT(6, "TCHAR:write\n");
+	device = (struct tape_device *) filp->private_data;
+	/* Find out block size and number of blocks */
+	if (device->char_data.block_size != 0) {
+		if (count < device->char_data.block_size) {
+			DBF_EVENT(3, "TCHAR:write smaller than block "
+				  "size was requested\n");
+			return -EINVAL;
+		}
+		block_size = device->char_data.block_size;
+		nblocks = count / block_size;
+	} else {
+		block_size = count;
+		nblocks = 1;
+	}
+
+	rc = tapechar_check_idalbuffer(device, block_size);
+	if (rc)
+		return rc;
+
+	DBF_EVENT(6,"TCHAR:nbytes: %lx\n", block_size);
+	DBF_EVENT(6, "TCHAR:nblocks: %x\n", nblocks);
+	/* Let the discipline build the ccw chain. */
+	request = device->discipline->write_block(device, block_size);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	rc = 0;
+	written = 0;
+	for (i = 0; i < nblocks; i++) {
+		/* Copy data from user space to idal buffer. */
+		if (idal_buffer_from_user(device->char_data.idal_buf,
+					  data, block_size)) {
+			rc = -EFAULT;
+			break;
+		}
+		rc = tape_do_io(device, request);
+		if (rc)
+			break;
+		DBF_EVENT(6, "TCHAR:wbytes: %lx\n",
+			  block_size - request->rescnt);
+		written += block_size - request->rescnt;
+		if (request->rescnt != 0)
+			break;
+		data += block_size;
+	}
+	tape_free_request(request);
+	if (rc == -ENOSPC) {
+		/*
+		 * Ok, the device has no more space. It has NOT written
+		 * the block.
+		 */
+		if (device->discipline->process_eov)
+			device->discipline->process_eov(device);
+		if (written > 0)
+			rc = 0;
+
+	}
+
+	/*
+	 * After doing a write we always need two tapemarks to correctly
+	 * terminate the tape (one to terminate the file, the second to
+	 * flag the end of recorded data.
+	 * Since process_eov positions the tape in front of the written
+	 * tapemark it doesn't hurt to write two marks again.
+	 */
+	if (!rc)
+		device->required_tapemarks = 2;
+
+	return rc ? rc : written;
+}
+
+/*
+ * Character frontend tape device open function.
+ */
+static int
+tapechar_open (struct inode *inode, struct file *filp)
+{
+	struct tape_device *device;
+	int minor, rc;
+
+	DBF_EVENT(6, "TCHAR:open: %i:%i\n",
+		imajor(file_inode(filp)),
+		iminor(file_inode(filp)));
+
+	if (imajor(file_inode(filp)) != tapechar_major)
+		return -ENODEV;
+
+	minor = iminor(file_inode(filp));
+	device = tape_find_device(minor / TAPE_MINORS_PER_DEV);
+	if (IS_ERR(device)) {
+		DBF_EVENT(3, "TCHAR:open: tape_find_device() failed\n");
+		return PTR_ERR(device);
+	}
+
+	rc = tape_open(device);
+	if (rc == 0) {
+		filp->private_data = device;
+		nonseekable_open(inode, filp);
+	} else
+		tape_put_device(device);
+
+	return rc;
+}
+
+/*
+ * Character frontend tape device release function.
+ */
+
+static int
+tapechar_release(struct inode *inode, struct file *filp)
+{
+	struct tape_device *device;
+
+	DBF_EVENT(6, "TCHAR:release: %x\n", iminor(inode));
+	device = (struct tape_device *) filp->private_data;
+
+	/*
+	 * If this is the rewinding tape minor then rewind. In that case we
+	 * write all required tapemarks. Otherwise only one to terminate the
+	 * file.
+	 */
+	if ((iminor(inode) & 1) != 0) {
+		if (device->required_tapemarks)
+			tape_std_terminate_write(device);
+		tape_mtop(device, MTREW, 1);
+	} else {
+		if (device->required_tapemarks > 1) {
+			if (tape_mtop(device, MTWEOF, 1) == 0)
+				device->required_tapemarks--;
+		}
+	}
+
+	if (device->char_data.idal_buf != NULL) {
+		idal_buffer_free(device->char_data.idal_buf);
+		device->char_data.idal_buf = NULL;
+	}
+	tape_release(device);
+	filp->private_data = NULL;
+	tape_put_device(device);
+
+	return 0;
+}
+
+/*
+ * Tape device io controls.
+ */
+static int
+__tapechar_ioctl(struct tape_device *device,
+		 unsigned int no, unsigned long data)
+{
+	int rc;
+
+	if (no == MTIOCTOP) {
+		struct mtop op;
+
+		if (copy_from_user(&op, (char __user *) data, sizeof(op)) != 0)
+			return -EFAULT;
+		if (op.mt_count < 0)
+			return -EINVAL;
+
+		/*
+		 * Operations that change tape position should write final
+		 * tapemarks.
+		 */
+		switch (op.mt_op) {
+			case MTFSF:
+			case MTBSF:
+			case MTFSR:
+			case MTBSR:
+			case MTREW:
+			case MTOFFL:
+			case MTEOM:
+			case MTRETEN:
+			case MTBSFM:
+			case MTFSFM:
+			case MTSEEK:
+				if (device->required_tapemarks)
+					tape_std_terminate_write(device);
+			default:
+				;
+		}
+		rc = tape_mtop(device, op.mt_op, op.mt_count);
+
+		if (op.mt_op == MTWEOF && rc == 0) {
+			if (op.mt_count > device->required_tapemarks)
+				device->required_tapemarks = 0;
+			else
+				device->required_tapemarks -= op.mt_count;
+		}
+		return rc;
+	}
+	if (no == MTIOCPOS) {
+		/* MTIOCPOS: query the tape position. */
+		struct mtpos pos;
+
+		rc = tape_mtop(device, MTTELL, 1);
+		if (rc < 0)
+			return rc;
+		pos.mt_blkno = rc;
+		if (copy_to_user((char __user *) data, &pos, sizeof(pos)) != 0)
+			return -EFAULT;
+		return 0;
+	}
+	if (no == MTIOCGET) {
+		/* MTIOCGET: query the tape drive status. */
+		struct mtget get;
+
+		memset(&get, 0, sizeof(get));
+		get.mt_type = MT_ISUNKNOWN;
+		get.mt_resid = 0 /* device->devstat.rescnt */;
+		get.mt_dsreg =
+			((device->char_data.block_size << MT_ST_BLKSIZE_SHIFT)
+			 & MT_ST_BLKSIZE_MASK);
+		/* FIXME: mt_gstat, mt_erreg, mt_fileno */
+		get.mt_gstat = 0;
+		get.mt_erreg = 0;
+		get.mt_fileno = 0;
+		get.mt_gstat  = device->tape_generic_status;
+
+		if (device->medium_state == MS_LOADED) {
+			rc = tape_mtop(device, MTTELL, 1);
+
+			if (rc < 0)
+				return rc;
+
+			if (rc == 0)
+				get.mt_gstat |= GMT_BOT(~0);
+
+			get.mt_blkno = rc;
+		}
+
+		if (copy_to_user((char __user *) data, &get, sizeof(get)) != 0)
+			return -EFAULT;
+
+		return 0;
+	}
+	/* Try the discipline ioctl function. */
+	if (device->discipline->ioctl_fn == NULL)
+		return -EINVAL;
+	return device->discipline->ioctl_fn(device, no, data);
+}
+
+static long
+tapechar_ioctl(struct file *filp, unsigned int no, unsigned long data)
+{
+	struct tape_device *device;
+	long rc;
+
+	DBF_EVENT(6, "TCHAR:ioct\n");
+
+	device = (struct tape_device *) filp->private_data;
+	mutex_lock(&device->mutex);
+	rc = __tapechar_ioctl(device, no, data);
+	mutex_unlock(&device->mutex);
+	return rc;
+}
+
+#ifdef CONFIG_COMPAT
+static long
+tapechar_compat_ioctl(struct file *filp, unsigned int no, unsigned long data)
+{
+	struct tape_device *device = filp->private_data;
+	int rval = -ENOIOCTLCMD;
+	unsigned long argp;
+
+	/* The 'arg' argument of any ioctl function may only be used for
+	 * pointers because of the compat pointer conversion.
+	 * Consider this when adding new ioctls.
+	 */
+	argp = (unsigned long) compat_ptr(data);
+	if (device->discipline->ioctl_fn) {
+		mutex_lock(&device->mutex);
+		rval = device->discipline->ioctl_fn(device, no, argp);
+		mutex_unlock(&device->mutex);
+		if (rval == -EINVAL)
+			rval = -ENOIOCTLCMD;
+	}
+
+	return rval;
+}
+#endif /* CONFIG_COMPAT */
+
+/*
+ * Initialize character device frontend.
+ */
+int
+tapechar_init (void)
+{
+	dev_t	dev;
+
+	if (alloc_chrdev_region(&dev, 0, 256, "tape") != 0)
+		return -1;
+
+	tapechar_major = MAJOR(dev);
+
+	return 0;
+}
+
+/*
+ * cleanup
+ */
+void
+tapechar_exit(void)
+{
+	unregister_chrdev_region(MKDEV(tapechar_major, 0), 256);
+}

drivers/s390/char/tape_class.c

@@ -0,0 +1,132 @@
+/*
+ * Copyright IBM Corp. 2004
+ *
+ * Tape class device support
+ *
+ * Author: Stefan Bader <shbader@de.ibm.com>
+ * Based on simple class device code by Greg K-H
+ */
+
+#define KMSG_COMPONENT "tape"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/slab.h>
+
+#include "tape_class.h"
+
+MODULE_AUTHOR("Stefan Bader <shbader@de.ibm.com>");
+MODULE_DESCRIPTION(
+	"Copyright IBM Corp. 2004   All Rights Reserved.\n"
+	"tape_class.c"
+);
+MODULE_LICENSE("GPL");
+
+static struct class *tape_class;
+
+/*
+ * Register a tape device and return a pointer to the cdev structure.
+ *
+ * device
+ *	The pointer to the struct device of the physical (base) device.
+ * drivername
+ *	The pointer to the drivers name for it's character devices.
+ * dev
+ *	The intended major/minor number. The major number may be 0 to
+ *	get a dynamic major number.
+ * fops
+ *	The pointer to the drivers file operations for the tape device.
+ * devname
+ *	The pointer to the name of the character device.
+ */
+struct tape_class_device *register_tape_dev(
+	struct device *		device,
+	dev_t			dev,
+	const struct file_operations *fops,
+	char *			device_name,
+	char *			mode_name)
+{
+	struct tape_class_device *	tcd;
+	int		rc;
+	char *		s;
+
+	tcd = kzalloc(sizeof(struct tape_class_device), GFP_KERNEL);
+	if (!tcd)
+		return ERR_PTR(-ENOMEM);
+
+	strncpy(tcd->device_name, device_name, TAPECLASS_NAME_LEN);
+	for (s = strchr(tcd->device_name, '/'); s; s = strchr(s, '/'))
+		*s = '!';
+	strncpy(tcd->mode_name, mode_name, TAPECLASS_NAME_LEN);
+	for (s = strchr(tcd->mode_name, '/'); s; s = strchr(s, '/'))
+		*s = '!';
+
+	tcd->char_device = cdev_alloc();
+	if (!tcd->char_device) {
+		rc = -ENOMEM;
+		goto fail_with_tcd;
+	}
+
+	tcd->char_device->owner = fops->owner;
+	tcd->char_device->ops   = fops;
+	tcd->char_device->dev   = dev;
+
+	rc = cdev_add(tcd->char_device, tcd->char_device->dev, 1);
+	if (rc)
+		goto fail_with_cdev;
+
+	tcd->class_device = device_create(tape_class, device,
+					  tcd->char_device->dev, NULL,
+					  "%s", tcd->device_name);
+	rc = PTR_RET(tcd->class_device);
+	if (rc)
+		goto fail_with_cdev;
+	rc = sysfs_create_link(
+		&device->kobj,
+		&tcd->class_device->kobj,
+		tcd->mode_name
+	);
+	if (rc)
+		goto fail_with_class_device;
+
+	return tcd;
+
+fail_with_class_device:
+	device_destroy(tape_class, tcd->char_device->dev);
+
+fail_with_cdev:
+	cdev_del(tcd->char_device);
+
+fail_with_tcd:
+	kfree(tcd);
+
+	return ERR_PTR(rc);
+}
+EXPORT_SYMBOL(register_tape_dev);
+
+void unregister_tape_dev(struct device *device, struct tape_class_device *tcd)
+{
+	if (tcd != NULL && !IS_ERR(tcd)) {
+		sysfs_remove_link(&device->kobj, tcd->mode_name);
+		device_destroy(tape_class, tcd->char_device->dev);
+		cdev_del(tcd->char_device);
+		kfree(tcd);
+	}
+}
+EXPORT_SYMBOL(unregister_tape_dev);
+
+
+static int __init tape_init(void)
+{
+	tape_class = class_create(THIS_MODULE, "tape390");
+
+	return 0;
+}
+
+static void __exit tape_exit(void)
+{
+	class_destroy(tape_class);
+	tape_class = NULL;
+}
+
+postcore_initcall(tape_init);
+module_exit(tape_exit);

drivers/s390/char/tape_class.h

@@ -0,0 +1,58 @@
+/*
+ * Copyright IBM Corp. 2004   All Rights Reserved.
+ *
+ * Tape class device support
+ *
+ * Author: Stefan Bader <shbader@de.ibm.com>
+ * Based on simple class device code by Greg K-H
+ */
+#ifndef __TAPE_CLASS_H__
+#define __TAPE_CLASS_H__
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/fs.h>
+#include <linux/major.h>
+#include <linux/cdev.h>
+
+#include <linux/device.h>
+#include <linux/kdev_t.h>
+
+#define TAPECLASS_NAME_LEN	32
+
+struct tape_class_device {
+	struct cdev		*char_device;
+	struct device		*class_device;
+	char			device_name[TAPECLASS_NAME_LEN];
+	char			mode_name[TAPECLASS_NAME_LEN];
+};
+
+/*
+ * Register a tape device and return a pointer to the tape class device
+ * created by the call.
+ *
+ * device
+ *	The pointer to the struct device of the physical (base) device.
+ * dev
+ *	The intended major/minor number. The major number may be 0 to
+ *	get a dynamic major number.
+ * fops
+ *	The pointer to the drivers file operations for the tape device.
+ * device_name
+ *	Pointer to the logical device name (will also be used as kobject name
+ *	of the cdev). This can also be called the name of the tape class
+ *	device.
+ * mode_name
+ *	Points to the name of the tape mode. This creates a link with that
+ *	name from the physical device to the logical device (class).
+ */
+struct tape_class_device *register_tape_dev(
+	struct device *		device,
+	dev_t			dev,
+	const struct file_operations *fops,
+	char *			device_name,
+	char *			node_name
+);
+void unregister_tape_dev(struct device *device, struct tape_class_device *tcd);
+
+#endif /* __TAPE_CLASS_H__ */

drivers/s390/char/tape_proc.c

@@ -0,0 +1,143 @@
+/*
+ *    tape device driver for S/390 and zSeries tapes.
+ *
+ *  S390 and zSeries version
+ *    Copyright IBM Corp. 2001
+ *    Author(s): Carsten Otte <cotte@de.ibm.com>
+ *		 Michael Holzheu <holzheu@de.ibm.com>
+ *		 Tuan Ngo-Anh <ngoanh@de.ibm.com>
+ *
+ * PROCFS Functions
+ */
+
+#define KMSG_COMPONENT "tape"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/module.h>
+#include <linux/vmalloc.h>
+#include <linux/seq_file.h>
+#include <linux/proc_fs.h>
+
+#define TAPE_DBF_AREA	tape_core_dbf
+
+#include "tape.h"
+
+static const char *tape_med_st_verbose[MS_SIZE] =
+{
+	[MS_UNKNOWN] = "UNKNOWN ",
+	[MS_LOADED] = "LOADED  ",
+	[MS_UNLOADED] = "UNLOADED"
+};
+
+/* our proc tapedevices entry */
+static struct proc_dir_entry *tape_proc_devices;
+
+/*
+ * Show function for /proc/tapedevices
+ */
+static int tape_proc_show(struct seq_file *m, void *v)
+{
+	struct tape_device *device;
+	struct tape_request *request;
+	const char *str;
+	unsigned long n;
+
+	n = (unsigned long) v - 1;
+	if (!n) {
+		seq_printf(m, "TapeNo\tBusID      CuType/Model\t"
+			"DevType/Model\tBlkSize\tState\tOp\tMedState\n");
+	}
+	device = tape_find_device(n);
+	if (IS_ERR(device))
+		return 0;
+	spin_lock_irq(get_ccwdev_lock(device->cdev));
+	seq_printf(m, "%d\t", (int) n);
+	seq_printf(m, "%-10.10s ", dev_name(&device->cdev->dev));
+	seq_printf(m, "%04X/", device->cdev->id.cu_type);
+	seq_printf(m, "%02X\t", device->cdev->id.cu_model);
+	seq_printf(m, "%04X/", device->cdev->id.dev_type);
+	seq_printf(m, "%02X\t\t", device->cdev->id.dev_model);
+	if (device->char_data.block_size == 0)
+		seq_printf(m, "auto\t");
+	else
+		seq_printf(m, "%i\t", device->char_data.block_size);
+	if (device->tape_state >= 0 &&
+	    device->tape_state < TS_SIZE)
+		str = tape_state_verbose[device->tape_state];
+	else
+		str = "UNKNOWN";
+	seq_printf(m, "%s\t", str);
+	if (!list_empty(&device->req_queue)) {
+		request = list_entry(device->req_queue.next,
+				     struct tape_request, list);
+		str = tape_op_verbose[request->op];
+	} else
+		str = "---";
+	seq_printf(m, "%s\t", str);
+	seq_printf(m, "%s\n", tape_med_st_verbose[device->medium_state]);
+	spin_unlock_irq(get_ccwdev_lock(device->cdev));
+	tape_put_device(device);
+        return 0;
+}
+
+static void *tape_proc_start(struct seq_file *m, loff_t *pos)
+{
+	if (*pos >= 256 / TAPE_MINORS_PER_DEV)
+		return NULL;
+	return (void *)((unsigned long) *pos + 1);
+}
+
+static void *tape_proc_next(struct seq_file *m, void *v, loff_t *pos)
+{
+	++*pos;
+	return tape_proc_start(m, pos);
+}
+
+static void tape_proc_stop(struct seq_file *m, void *v)
+{
+}
+
+static const struct seq_operations tape_proc_seq = {
+	.start		= tape_proc_start,
+	.next		= tape_proc_next,
+	.stop		= tape_proc_stop,
+	.show		= tape_proc_show,
+};
+
+static int tape_proc_open(struct inode *inode, struct file *file)
+{
+	return seq_open(file, &tape_proc_seq);
+}
+
+static const struct file_operations tape_proc_ops =
+{
+	.owner		= THIS_MODULE,
+	.open		= tape_proc_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= seq_release,
+};
+
+/*
+ * Initialize procfs stuff on startup
+ */
+void
+tape_proc_init(void)
+{
+	tape_proc_devices =
+		proc_create("tapedevices", S_IFREG | S_IRUGO | S_IWUSR, NULL,
+			    &tape_proc_ops);
+	if (tape_proc_devices == NULL) {
+		return;
+	}
+}
+
+/*
+ * Cleanup all stuff registered to the procfs
+ */
+void
+tape_proc_cleanup(void)
+{
+	if (tape_proc_devices != NULL)
+		remove_proc_entry ("tapedevices", NULL);
+}

drivers/s390/char/tape_std.c

@@ -0,0 +1,750 @@
+/*
+ *    standard tape device functions for ibm tapes.
+ *
+ *  S390 and zSeries version
+ *    Copyright IBM Corp. 2001, 2002
+ *    Author(s): Carsten Otte <cotte@de.ibm.com>
+ *		 Michael Holzheu <holzheu@de.ibm.com>
+ *		 Tuan Ngo-Anh <ngoanh@de.ibm.com>
+ *		 Martin Schwidefsky <schwidefsky@de.ibm.com>
+ *		 Stefan Bader <shbader@de.ibm.com>
+ */
+
+#define KMSG_COMPONENT "tape"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/stddef.h>
+#include <linux/kernel.h>
+#include <linux/bio.h>
+#include <linux/timer.h>
+
+#include <asm/types.h>
+#include <asm/idals.h>
+#include <asm/ebcdic.h>
+#include <asm/tape390.h>
+
+#define TAPE_DBF_AREA	tape_core_dbf
+
+#include "tape.h"
+#include "tape_std.h"
+
+/*
+ * tape_std_assign
+ */
+static void
+tape_std_assign_timeout(unsigned long data)
+{
+	struct tape_request *	request;
+	struct tape_device *	device;
+	int rc;
+
+	request = (struct tape_request *) data;
+	device = request->device;
+	BUG_ON(!device);
+
+	DBF_EVENT(3, "%08x: Assignment timeout. Device busy.\n",
+			device->cdev_id);
+	rc = tape_cancel_io(device, request);
+	if(rc)
+		DBF_EVENT(3, "(%08x): Assign timeout: Cancel failed with rc = "
+			  "%i\n", device->cdev_id, rc);
+}
+
+int
+tape_std_assign(struct tape_device *device)
+{
+	int                  rc;
+	struct timer_list    timeout;
+	struct tape_request *request;
+
+	request = tape_alloc_request(2, 11);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+
+	request->op = TO_ASSIGN;
+	tape_ccw_cc(request->cpaddr, ASSIGN, 11, request->cpdata);
+	tape_ccw_end(request->cpaddr + 1, NOP, 0, NULL);
+
+	/*
+	 * The assign command sometimes blocks if the device is assigned
+	 * to another host (actually this shouldn't happen but it does).
+	 * So we set up a timeout for this call.
+	 */
+	init_timer_on_stack(&timeout);
+	timeout.function = tape_std_assign_timeout;
+	timeout.data     = (unsigned long) request;
+	timeout.expires  = jiffies + 2 * HZ;
+	add_timer(&timeout);
+
+	rc = tape_do_io_interruptible(device, request);
+
+	del_timer_sync(&timeout);
+	destroy_timer_on_stack(&timeout);
+
+	if (rc != 0) {
+		DBF_EVENT(3, "%08x: assign failed - device might be busy\n",
+			device->cdev_id);
+	} else {
+		DBF_EVENT(3, "%08x: Tape assigned\n", device->cdev_id);
+	}
+	tape_free_request(request);
+	return rc;
+}
+
+/*
+ * tape_std_unassign
+ */
+int
+tape_std_unassign (struct tape_device *device)
+{
+	int                  rc;
+	struct tape_request *request;
+
+	if (device->tape_state == TS_NOT_OPER) {
+		DBF_EVENT(3, "(%08x): Can't unassign device\n",
+			device->cdev_id);
+		return -EIO;
+	}
+
+	request = tape_alloc_request(2, 11);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+
+	request->op = TO_UNASSIGN;
+	tape_ccw_cc(request->cpaddr, UNASSIGN, 11, request->cpdata);
+	tape_ccw_end(request->cpaddr + 1, NOP, 0, NULL);
+
+	if ((rc = tape_do_io(device, request)) != 0) {
+		DBF_EVENT(3, "%08x: Unassign failed\n", device->cdev_id);
+	} else {
+		DBF_EVENT(3, "%08x: Tape unassigned\n", device->cdev_id);
+	}
+	tape_free_request(request);
+	return rc;
+}
+
+/*
+ * TAPE390_DISPLAY: Show a string on the tape display.
+ */
+int
+tape_std_display(struct tape_device *device, struct display_struct *disp)
+{
+	struct tape_request *request;
+	int rc;
+
+	request = tape_alloc_request(2, 17);
+	if (IS_ERR(request)) {
+		DBF_EVENT(3, "TAPE: load display failed\n");
+		return PTR_ERR(request);
+	}
+	request->op = TO_DIS;
+
+	*(unsigned char *) request->cpdata = disp->cntrl;
+	DBF_EVENT(5, "TAPE: display cntrl=%04x\n", disp->cntrl);
+	memcpy(((unsigned char *) request->cpdata) + 1, disp->message1, 8);
+	memcpy(((unsigned char *) request->cpdata) + 9, disp->message2, 8);
+	ASCEBC(((unsigned char*) request->cpdata) + 1, 16);
+
+	tape_ccw_cc(request->cpaddr, LOAD_DISPLAY, 17, request->cpdata);
+	tape_ccw_end(request->cpaddr + 1, NOP, 0, NULL);
+
+	rc = tape_do_io_interruptible(device, request);
+	tape_free_request(request);
+	return rc;
+}
+
+/*
+ * Read block id.
+ */
+int
+tape_std_read_block_id(struct tape_device *device, __u64 *id)
+{
+	struct tape_request *request;
+	int rc;
+
+	request = tape_alloc_request(3, 8);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	request->op = TO_RBI;
+	/* setup ccws */
+	tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1, device->modeset_byte);
+	tape_ccw_cc(request->cpaddr + 1, READ_BLOCK_ID, 8, request->cpdata);
+	tape_ccw_end(request->cpaddr + 2, NOP, 0, NULL);
+	/* execute it */
+	rc = tape_do_io(device, request);
+	if (rc == 0)
+		/* Get result from read buffer. */
+		*id = *(__u64 *) request->cpdata;
+	tape_free_request(request);
+	return rc;
+}
+
+int
+tape_std_terminate_write(struct tape_device *device)
+{
+	int rc;
+
+	if(device->required_tapemarks == 0)
+		return 0;
+
+	DBF_LH(5, "tape%d: terminate write %dxEOF\n", device->first_minor,
+		device->required_tapemarks);
+
+	rc = tape_mtop(device, MTWEOF, device->required_tapemarks);
+	if (rc)
+		return rc;
+
+	device->required_tapemarks = 0;
+	return tape_mtop(device, MTBSR, 1);
+}
+
+/*
+ * MTLOAD: Loads the tape.
+ * The default implementation just wait until the tape medium state changes
+ * to MS_LOADED.
+ */
+int
+tape_std_mtload(struct tape_device *device, int count)
+{
+	return wait_event_interruptible(device->state_change_wq,
+		(device->medium_state == MS_LOADED));
+}
+
+/*
+ * MTSETBLK: Set block size.
+ */
+int
+tape_std_mtsetblk(struct tape_device *device, int count)
+{
+	struct idal_buffer *new;
+
+	DBF_LH(6, "tape_std_mtsetblk(%d)\n", count);
+	if (count <= 0) {
+		/*
+		 * Just set block_size to 0. tapechar_read/tapechar_write
+		 * will realloc the idal buffer if a bigger one than the
+		 * current is needed.
+		 */
+		device->char_data.block_size = 0;
+		return 0;
+	}
+	if (device->char_data.idal_buf != NULL &&
+	    device->char_data.idal_buf->size == count)
+		/* We already have a idal buffer of that size. */
+		return 0;
+
+	if (count > MAX_BLOCKSIZE) {
+		DBF_EVENT(3, "Invalid block size (%d > %d) given.\n",
+			count, MAX_BLOCKSIZE);
+		return -EINVAL;
+	}
+
+	/* Allocate a new idal buffer. */
+	new = idal_buffer_alloc(count, 0);
+	if (IS_ERR(new))
+		return -ENOMEM;
+	if (device->char_data.idal_buf != NULL)
+		idal_buffer_free(device->char_data.idal_buf);
+	device->char_data.idal_buf = new;
+	device->char_data.block_size = count;
+
+	DBF_LH(6, "new blocksize is %d\n", device->char_data.block_size);
+
+	return 0;
+}
+
+/*
+ * MTRESET: Set block size to 0.
+ */
+int
+tape_std_mtreset(struct tape_device *device, int count)
+{
+	DBF_EVENT(6, "TCHAR:devreset:\n");
+	device->char_data.block_size = 0;
+	return 0;
+}
+
+/*
+ * MTFSF: Forward space over 'count' file marks. The tape is positioned
+ * at the EOT (End of Tape) side of the file mark.
+ */
+int
+tape_std_mtfsf(struct tape_device *device, int mt_count)
+{
+	struct tape_request *request;
+	struct ccw1 *ccw;
+
+	request = tape_alloc_request(mt_count + 2, 0);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	request->op = TO_FSF;
+	/* setup ccws */
+	ccw = tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1,
+			  device->modeset_byte);
+	ccw = tape_ccw_repeat(ccw, FORSPACEFILE, mt_count);
+	ccw = tape_ccw_end(ccw, NOP, 0, NULL);
+
+	/* execute it */
+	return tape_do_io_free(device, request);
+}
+
+/*
+ * MTFSR: Forward space over 'count' tape blocks (blocksize is set
+ * via MTSETBLK.
+ */
+int
+tape_std_mtfsr(struct tape_device *device, int mt_count)
+{
+	struct tape_request *request;
+	struct ccw1 *ccw;
+	int rc;
+
+	request = tape_alloc_request(mt_count + 2, 0);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	request->op = TO_FSB;
+	/* setup ccws */
+	ccw = tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1,
+			  device->modeset_byte);
+	ccw = tape_ccw_repeat(ccw, FORSPACEBLOCK, mt_count);
+	ccw = tape_ccw_end(ccw, NOP, 0, NULL);
+
+	/* execute it */
+	rc = tape_do_io(device, request);
+	if (rc == 0 && request->rescnt > 0) {
+		DBF_LH(3, "FSR over tapemark\n");
+		rc = 1;
+	}
+	tape_free_request(request);
+
+	return rc;
+}
+
+/*
+ * MTBSR: Backward space over 'count' tape blocks.
+ * (blocksize is set via MTSETBLK.
+ */
+int
+tape_std_mtbsr(struct tape_device *device, int mt_count)
+{
+	struct tape_request *request;
+	struct ccw1 *ccw;
+	int rc;
+
+	request = tape_alloc_request(mt_count + 2, 0);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	request->op = TO_BSB;
+	/* setup ccws */
+	ccw = tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1,
+			  device->modeset_byte);
+	ccw = tape_ccw_repeat(ccw, BACKSPACEBLOCK, mt_count);
+	ccw = tape_ccw_end(ccw, NOP, 0, NULL);
+
+	/* execute it */
+	rc = tape_do_io(device, request);
+	if (rc == 0 && request->rescnt > 0) {
+		DBF_LH(3, "BSR over tapemark\n");
+		rc = 1;
+	}
+	tape_free_request(request);
+
+	return rc;
+}
+
+/*
+ * MTWEOF: Write 'count' file marks at the current position.
+ */
+int
+tape_std_mtweof(struct tape_device *device, int mt_count)
+{
+	struct tape_request *request;
+	struct ccw1 *ccw;
+
+	request = tape_alloc_request(mt_count + 2, 0);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	request->op = TO_WTM;
+	/* setup ccws */
+	ccw = tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1,
+			  device->modeset_byte);
+	ccw = tape_ccw_repeat(ccw, WRITETAPEMARK, mt_count);
+	ccw = tape_ccw_end(ccw, NOP, 0, NULL);
+
+	/* execute it */
+	return tape_do_io_free(device, request);
+}
+
+/*
+ * MTBSFM: Backward space over 'count' file marks.
+ * The tape is positioned at the BOT (Begin Of Tape) side of the
+ * last skipped file mark.
+ */
+int
+tape_std_mtbsfm(struct tape_device *device, int mt_count)
+{
+	struct tape_request *request;
+	struct ccw1 *ccw;
+
+	request = tape_alloc_request(mt_count + 2, 0);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	request->op = TO_BSF;
+	/* setup ccws */
+	ccw = tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1,
+			  device->modeset_byte);
+	ccw = tape_ccw_repeat(ccw, BACKSPACEFILE, mt_count);
+	ccw = tape_ccw_end(ccw, NOP, 0, NULL);
+
+	/* execute it */
+	return tape_do_io_free(device, request);
+}
+
+/*
+ * MTBSF: Backward space over 'count' file marks. The tape is positioned at
+ * the EOT (End of Tape) side of the last skipped file mark.
+ */
+int
+tape_std_mtbsf(struct tape_device *device, int mt_count)
+{
+	struct tape_request *request;
+	struct ccw1 *ccw;
+	int rc;
+
+	request = tape_alloc_request(mt_count + 2, 0);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	request->op = TO_BSF;
+	/* setup ccws */
+	ccw = tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1,
+			  device->modeset_byte);
+	ccw = tape_ccw_repeat(ccw, BACKSPACEFILE, mt_count);
+	ccw = tape_ccw_end(ccw, NOP, 0, NULL);
+	/* execute it */
+	rc = tape_do_io_free(device, request);
+	if (rc == 0) {
+		rc = tape_mtop(device, MTFSR, 1);
+		if (rc > 0)
+			rc = 0;
+	}
+	return rc;
+}
+
+/*
+ * MTFSFM: Forward space over 'count' file marks.
+ * The tape is positioned at the BOT (Begin Of Tape) side
+ * of the last skipped file mark.
+ */
+int
+tape_std_mtfsfm(struct tape_device *device, int mt_count)
+{
+	struct tape_request *request;
+	struct ccw1 *ccw;
+	int rc;
+
+	request = tape_alloc_request(mt_count + 2, 0);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	request->op = TO_FSF;
+	/* setup ccws */
+	ccw = tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1,
+			  device->modeset_byte);
+	ccw = tape_ccw_repeat(ccw, FORSPACEFILE, mt_count);
+	ccw = tape_ccw_end(ccw, NOP, 0, NULL);
+	/* execute it */
+	rc = tape_do_io_free(device, request);
+	if (rc == 0) {
+		rc = tape_mtop(device, MTBSR, 1);
+		if (rc > 0)
+			rc = 0;
+	}
+
+	return rc;
+}
+
+/*
+ * MTREW: Rewind the tape.
+ */
+int
+tape_std_mtrew(struct tape_device *device, int mt_count)
+{
+	struct tape_request *request;
+
+	request = tape_alloc_request(3, 0);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	request->op = TO_REW;
+	/* setup ccws */
+	tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1,
+		    device->modeset_byte);
+	tape_ccw_cc(request->cpaddr + 1, REWIND, 0, NULL);
+	tape_ccw_end(request->cpaddr + 2, NOP, 0, NULL);
+
+	/* execute it */
+	return tape_do_io_free(device, request);
+}
+
+/*
+ * MTOFFL: Rewind the tape and put the drive off-line.
+ * Implement 'rewind unload'
+ */
+int
+tape_std_mtoffl(struct tape_device *device, int mt_count)
+{
+	struct tape_request *request;
+
+	request = tape_alloc_request(3, 0);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	request->op = TO_RUN;
+	/* setup ccws */
+	tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1, device->modeset_byte);
+	tape_ccw_cc(request->cpaddr + 1, REWIND_UNLOAD, 0, NULL);
+	tape_ccw_end(request->cpaddr + 2, NOP, 0, NULL);
+
+	/* execute it */
+	return tape_do_io_free(device, request);
+}
+
+/*
+ * MTNOP: 'No operation'.
+ */
+int
+tape_std_mtnop(struct tape_device *device, int mt_count)
+{
+	struct tape_request *request;
+
+	request = tape_alloc_request(2, 0);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	request->op = TO_NOP;
+	/* setup ccws */
+	tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1, device->modeset_byte);
+	tape_ccw_end(request->cpaddr + 1, NOP, 0, NULL);
+	/* execute it */
+	return tape_do_io_free(device, request);
+}
+
+/*
+ * MTEOM: positions at the end of the portion of the tape already used
+ * for recordind data. MTEOM positions after the last file mark, ready for
+ * appending another file.
+ */
+int
+tape_std_mteom(struct tape_device *device, int mt_count)
+{
+	int rc;
+
+	/*
+	 * Seek from the beginning of tape (rewind).
+	 */
+	if ((rc = tape_mtop(device, MTREW, 1)) < 0)
+		return rc;
+
+	/*
+	 * The logical end of volume is given by two sewuential tapemarks.
+	 * Look for this by skipping to the next file (over one tapemark)
+	 * and then test for another one (fsr returns 1 if a tapemark was
+	 * encountered).
+	 */
+	do {
+		if ((rc = tape_mtop(device, MTFSF, 1)) < 0)
+			return rc;
+		if ((rc = tape_mtop(device, MTFSR, 1)) < 0)
+			return rc;
+	} while (rc == 0);
+
+	return tape_mtop(device, MTBSR, 1);
+}
+
+/*
+ * MTRETEN: Retension the tape, i.e. forward space to end of tape and rewind.
+ */
+int
+tape_std_mtreten(struct tape_device *device, int mt_count)
+{
+	struct tape_request *request;
+
+	request = tape_alloc_request(4, 0);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	request->op = TO_FSF;
+	/* setup ccws */
+	tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1, device->modeset_byte);
+	tape_ccw_cc(request->cpaddr + 1,FORSPACEFILE, 0, NULL);
+	tape_ccw_cc(request->cpaddr + 2, NOP, 0, NULL);
+	tape_ccw_end(request->cpaddr + 3, CCW_CMD_TIC, 0, request->cpaddr);
+	/* execute it, MTRETEN rc gets ignored */
+	tape_do_io_interruptible(device, request);
+	tape_free_request(request);
+	return tape_mtop(device, MTREW, 1);
+}
+
+/*
+ * MTERASE: erases the tape.
+ */
+int
+tape_std_mterase(struct tape_device *device, int mt_count)
+{
+	struct tape_request *request;
+
+	request = tape_alloc_request(6, 0);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	request->op = TO_DSE;
+	/* setup ccws */
+	tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1, device->modeset_byte);
+	tape_ccw_cc(request->cpaddr + 1, REWIND, 0, NULL);
+	tape_ccw_cc(request->cpaddr + 2, ERASE_GAP, 0, NULL);
+	tape_ccw_cc(request->cpaddr + 3, DATA_SEC_ERASE, 0, NULL);
+	tape_ccw_cc(request->cpaddr + 4, REWIND, 0, NULL);
+	tape_ccw_end(request->cpaddr + 5, NOP, 0, NULL);
+
+	/* execute it */
+	return tape_do_io_free(device, request);
+}
+
+/*
+ * MTUNLOAD: Rewind the tape and unload it.
+ */
+int
+tape_std_mtunload(struct tape_device *device, int mt_count)
+{
+	return tape_mtop(device, MTOFFL, mt_count);
+}
+
+/*
+ * MTCOMPRESSION: used to enable compression.
+ * Sets the IDRC on/off.
+ */
+int
+tape_std_mtcompression(struct tape_device *device, int mt_count)
+{
+	struct tape_request *request;
+
+	if (mt_count < 0 || mt_count > 1) {
+		DBF_EXCEPTION(6, "xcom parm\n");
+		return -EINVAL;
+	}
+	request = tape_alloc_request(2, 0);
+	if (IS_ERR(request))
+		return PTR_ERR(request);
+	request->op = TO_NOP;
+	/* setup ccws */
+	if (mt_count == 0)
+		*device->modeset_byte &= ~0x08;
+	else
+		*device->modeset_byte |= 0x08;
+	tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1, device->modeset_byte);
+	tape_ccw_end(request->cpaddr + 1, NOP, 0, NULL);
+	/* execute it */
+	return tape_do_io_free(device, request);
+}
+
+/*
+ * Read Block
+ */
+struct tape_request *
+tape_std_read_block(struct tape_device *device, size_t count)
+{
+	struct tape_request *request;
+
+	/*
+	 * We have to alloc 4 ccws in order to be able to transform request
+	 * into a read backward request in error case.
+	 */
+	request = tape_alloc_request(4, 0);
+	if (IS_ERR(request)) {
+		DBF_EXCEPTION(6, "xrbl fail");
+		return request;
+	}
+	request->op = TO_RFO;
+	tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1, device->modeset_byte);
+	tape_ccw_end_idal(request->cpaddr + 1, READ_FORWARD,
+			  device->char_data.idal_buf);
+	DBF_EVENT(6, "xrbl ccwg\n");
+	return request;
+}
+
+/*
+ * Read Block backward transformation function.
+ */
+void
+tape_std_read_backward(struct tape_device *device, struct tape_request *request)
+{
+	/*
+	 * We have allocated 4 ccws in tape_std_read, so we can now
+	 * transform the request to a read backward, followed by a
+	 * forward space block.
+	 */
+	request->op = TO_RBA;
+	tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1, device->modeset_byte);
+	tape_ccw_cc_idal(request->cpaddr + 1, READ_BACKWARD,
+			 device->char_data.idal_buf);
+	tape_ccw_cc(request->cpaddr + 2, FORSPACEBLOCK, 0, NULL);
+	tape_ccw_end(request->cpaddr + 3, NOP, 0, NULL);
+	DBF_EVENT(6, "xrop ccwg");}
+
+/*
+ * Write Block
+ */
+struct tape_request *
+tape_std_write_block(struct tape_device *device, size_t count)
+{
+	struct tape_request *request;
+
+	request = tape_alloc_request(2, 0);
+	if (IS_ERR(request)) {
+		DBF_EXCEPTION(6, "xwbl fail\n");
+		return request;
+	}
+	request->op = TO_WRI;
+	tape_ccw_cc(request->cpaddr, MODE_SET_DB, 1, device->modeset_byte);
+	tape_ccw_end_idal(request->cpaddr + 1, WRITE_CMD,
+			  device->char_data.idal_buf);
+	DBF_EVENT(6, "xwbl ccwg\n");
+	return request;
+}
+
+/*
+ * This routine is called by frontend after an ENOSP on write
+ */
+void
+tape_std_process_eov(struct tape_device *device)
+{
+	/*
+	 * End of volume: We have to backspace the last written record, then
+	 * we TRY to write a tapemark and then backspace over the written TM
+	 */
+	if (tape_mtop(device, MTBSR, 1) == 0 &&
+	    tape_mtop(device, MTWEOF, 1) == 0) {
+		tape_mtop(device, MTBSR, 1);
+	}
+}
+
+EXPORT_SYMBOL(tape_std_assign);
+EXPORT_SYMBOL(tape_std_unassign);
+EXPORT_SYMBOL(tape_std_display);
+EXPORT_SYMBOL(tape_std_read_block_id);
+EXPORT_SYMBOL(tape_std_mtload);
+EXPORT_SYMBOL(tape_std_mtsetblk);
+EXPORT_SYMBOL(tape_std_mtreset);
+EXPORT_SYMBOL(tape_std_mtfsf);
+EXPORT_SYMBOL(tape_std_mtfsr);
+EXPORT_SYMBOL(tape_std_mtbsr);
+EXPORT_SYMBOL(tape_std_mtweof);
+EXPORT_SYMBOL(tape_std_mtbsfm);
+EXPORT_SYMBOL(tape_std_mtbsf);
+EXPORT_SYMBOL(tape_std_mtfsfm);
+EXPORT_SYMBOL(tape_std_mtrew);
+EXPORT_SYMBOL(tape_std_mtoffl);
+EXPORT_SYMBOL(tape_std_mtnop);
+EXPORT_SYMBOL(tape_std_mteom);
+EXPORT_SYMBOL(tape_std_mtreten);
+EXPORT_SYMBOL(tape_std_mterase);
+EXPORT_SYMBOL(tape_std_mtunload);
+EXPORT_SYMBOL(tape_std_mtcompression);
+EXPORT_SYMBOL(tape_std_read_block);
+EXPORT_SYMBOL(tape_std_read_backward);
+EXPORT_SYMBOL(tape_std_write_block);
+EXPORT_SYMBOL(tape_std_process_eov);

drivers/s390/char/tape_std.h

@@ -0,0 +1,154 @@
+/*
+ *    standard tape device functions for ibm tapes.
+ *
+ *    Copyright IBM Corp. 2001, 2006
+ *    Author(s): Carsten Otte <cotte@de.ibm.com>
+ *		 Tuan Ngo-Anh <ngoanh@de.ibm.com>
+ *		 Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#ifndef _TAPE_STD_H
+#define _TAPE_STD_H
+
+#include <asm/tape390.h>
+
+/*
+ * Biggest block size to handle. Currently 64K because we only build
+ * channel programs without data chaining.
+ */
+#define MAX_BLOCKSIZE   65535
+
+/*
+ * The CCW commands for the Tape type of command.
+ */
+#define INVALID_00		0x00	/* Invalid cmd */
+#define BACKSPACEBLOCK		0x27	/* Back Space block */
+#define BACKSPACEFILE		0x2f	/* Back Space file */
+#define DATA_SEC_ERASE		0x97	/* Data security erase */
+#define ERASE_GAP		0x17	/* Erase Gap */
+#define FORSPACEBLOCK		0x37	/* Forward space block */
+#define FORSPACEFILE		0x3F	/* Forward Space file */
+#define FORCE_STREAM_CNT	0xEB	/* Forced streaming count # */
+#define NOP			0x03	/* No operation	*/
+#define READ_FORWARD		0x02	/* Read forward */
+#define REWIND			0x07	/* Rewind */
+#define REWIND_UNLOAD		0x0F	/* Rewind and Unload */
+#define SENSE			0x04	/* Sense */
+#define NEW_MODE_SET		0xEB	/* Guess it is Mode set */
+#define WRITE_CMD		0x01	/* Write */
+#define WRITETAPEMARK		0x1F	/* Write Tape Mark */
+
+#define ASSIGN			0xB7	/* 3420 REJECT,3480 OK	*/
+#define CONTROL_ACCESS		0xE3	/* Set high speed */
+#define DIAG_MODE_SET		0x0B	/* 3420 NOP, 3480 REJECT */
+#define LOAD_DISPLAY		0x9F	/* 3420 REJECT,3480 OK */
+#define LOCATE			0x4F	/* 3420 REJ, 3480 NOP */
+#define LOOP_WRITE_TO_READ	0x8B	/* 3480 REJECT */
+#define MODE_SET_DB		0xDB	/* 3420 REJECT,3480 OK */
+#define MODE_SET_C3		0xC3	/* for 3420 */
+#define MODE_SET_CB		0xCB	/* for 3420 */
+#define MODE_SET_D3		0xD3	/* for 3420 */
+#define READ_BACKWARD		0x0C	/* */
+#define READ_BLOCK_ID		0x22	/* 3420 REJECT,3480 OK */
+#define READ_BUFFER		0x12	/* 3420 REJECT,3480 OK */
+#define READ_BUFF_LOG		0x24	/* 3420 REJECT,3480 OK */
+#define RELEASE			0xD4	/* 3420 NOP, 3480 REJECT */
+#define REQ_TRK_IN_ERROR	0x1B	/* 3420 NOP, 3480 REJECT */
+#define RESERVE			0xF4	/* 3420 NOP, 3480 REJECT */
+#define SENSE_GROUP_ID		0x34	/* 3420 REJECT,3480 OK */
+#define SENSE_ID		0xE4	/* 3420 REJECT,3480 OK */
+#define READ_DEV_CHAR		0x64	/* Read device characteristics */
+#define SET_DIAGNOSE		0x4B	/* 3420 NOP, 3480 REJECT */
+#define SET_GROUP_ID		0xAF	/* 3420 REJECT,3480 OK */
+#define SET_TAPE_WRITE_IMMED	0xC3	/* for 3480 */
+#define SUSPEND			0x5B	/* 3420 REJ, 3480 NOP */
+#define SYNC			0x43	/* Synchronize (flush buffer) */
+#define UNASSIGN		0xC7	/* 3420 REJECT,3480 OK */
+#define PERF_SUBSYS_FUNC	0x77	/* 3490 CMD */
+#define READ_CONFIG_DATA	0xFA	/* 3490 CMD */
+#define READ_MESSAGE_ID		0x4E	/* 3490 CMD */
+#define READ_SUBSYS_DATA	0x3E	/* 3490 CMD */
+#define SET_INTERFACE_ID	0x73	/* 3490 CMD */
+
+#define SENSE_COMMAND_REJECT		0x80
+#define SENSE_INTERVENTION_REQUIRED	0x40
+#define SENSE_BUS_OUT_CHECK		0x20
+#define SENSE_EQUIPMENT_CHECK		0x10
+#define SENSE_DATA_CHECK		0x08
+#define SENSE_OVERRUN			0x04
+#define SENSE_DEFERRED_UNIT_CHECK	0x02
+#define SENSE_ASSIGNED_ELSEWHERE	0x01
+
+#define SENSE_LOCATE_FAILURE		0x80
+#define SENSE_DRIVE_ONLINE		0x40
+#define SENSE_RESERVED			0x20
+#define SENSE_RECORD_SEQUENCE_ERR	0x10
+#define SENSE_BEGINNING_OF_TAPE		0x08
+#define SENSE_WRITE_MODE		0x04
+#define SENSE_WRITE_PROTECT		0x02
+#define SENSE_NOT_CAPABLE		0x01
+
+#define SENSE_CHANNEL_ADAPTER_CODE	0xE0
+#define SENSE_CHANNEL_ADAPTER_LOC	0x10
+#define SENSE_REPORTING_CU		0x08
+#define SENSE_AUTOMATIC_LOADER		0x04
+#define SENSE_TAPE_SYNC_MODE		0x02
+#define SENSE_TAPE_POSITIONING		0x01
+
+/* discipline functions */
+struct tape_request *tape_std_read_block(struct tape_device *, size_t);
+void tape_std_read_backward(struct tape_device *device,
+			    struct tape_request *request);
+struct tape_request *tape_std_write_block(struct tape_device *, size_t);
+void tape_std_check_locate(struct tape_device *, struct tape_request *);
+
+/* Some non-mtop commands. */
+int tape_std_assign(struct tape_device *);
+int tape_std_unassign(struct tape_device *);
+int tape_std_read_block_id(struct tape_device *device, __u64 *id);
+int tape_std_display(struct tape_device *, struct display_struct *disp);
+int tape_std_terminate_write(struct tape_device *);
+
+/* Standard magnetic tape commands. */
+int tape_std_mtbsf(struct tape_device *, int);
+int tape_std_mtbsfm(struct tape_device *, int);
+int tape_std_mtbsr(struct tape_device *, int);
+int tape_std_mtcompression(struct tape_device *, int);
+int tape_std_mteom(struct tape_device *, int);
+int tape_std_mterase(struct tape_device *, int);
+int tape_std_mtfsf(struct tape_device *, int);
+int tape_std_mtfsfm(struct tape_device *, int);
+int tape_std_mtfsr(struct tape_device *, int);
+int tape_std_mtload(struct tape_device *, int);
+int tape_std_mtnop(struct tape_device *, int);
+int tape_std_mtoffl(struct tape_device *, int);
+int tape_std_mtreset(struct tape_device *, int);
+int tape_std_mtreten(struct tape_device *, int);
+int tape_std_mtrew(struct tape_device *, int);
+int tape_std_mtsetblk(struct tape_device *, int);
+int tape_std_mtunload(struct tape_device *, int);
+int tape_std_mtweof(struct tape_device *, int);
+
+/* Event handlers */
+void tape_std_default_handler(struct tape_device *);
+void tape_std_unexpect_uchk_handler(struct tape_device *);
+void tape_std_irq(struct tape_device *);
+void tape_std_process_eov(struct tape_device *);
+
+// the error recovery stuff:
+void tape_std_error_recovery(struct tape_device *);
+void tape_std_error_recovery_has_failed(struct tape_device *,int error_id);
+void tape_std_error_recovery_succeded(struct tape_device *);
+void tape_std_error_recovery_do_retry(struct tape_device *);
+void tape_std_error_recovery_read_opposite(struct tape_device *);
+void tape_std_error_recovery_HWBUG(struct tape_device *, int condno);
+
+/* S390 tape types */
+enum s390_tape_type {
+        tape_3480,
+        tape_3490,
+        tape_3590,
+        tape_3592,
+};
+
+#endif // _TAPE_STD_H

drivers/s390/char/tty3270.h

@@ -0,0 +1,14 @@
+/*
+ *    Copyright IBM Corp. 2007
+ *
+ */
+
+#ifndef __DRIVERS_S390_CHAR_TTY3270_H
+#define __DRIVERS_S390_CHAR_TTY3270_H
+
+#include <linux/tty.h>
+#include <linux/tty_driver.h>
+
+extern struct tty_driver *tty3270_driver;
+
+#endif /* __DRIVERS_S390_CHAR_TTY3270_H */

drivers/s390/char/vmcp.c

@@ -0,0 +1,213 @@
+/*
+ * Copyright IBM Corp. 2004, 2010
+ * Interface implementation for communication with the z/VM control program
+ *
+ * Author(s): Christian Borntraeger <borntraeger@de.ibm.com>
+ *
+ * z/VMs CP offers the possibility to issue commands via the diagnose code 8
+ * this driver implements a character device that issues these commands and
+ * returns the answer of CP.
+ *
+ * The idea of this driver is based on cpint from Neale Ferguson and #CP in CMS
+ */
+
+#include <linux/fs.h>
+#include <linux/init.h>
+#include <linux/compat.h>
+#include <linux/kernel.h>
+#include <linux/miscdevice.h>
+#include <linux/slab.h>
+#include <linux/export.h>
+#include <asm/compat.h>
+#include <asm/cpcmd.h>
+#include <asm/debug.h>
+#include <asm/uaccess.h>
+#include "vmcp.h"
+
+static debug_info_t *vmcp_debug;
+
+static int vmcp_open(struct inode *inode, struct file *file)
+{
+	struct vmcp_session *session;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	session = kmalloc(sizeof(*session), GFP_KERNEL);
+	if (!session)
+		return -ENOMEM;
+
+	session->bufsize = PAGE_SIZE;
+	session->response = NULL;
+	session->resp_size = 0;
+	mutex_init(&session->mutex);
+	file->private_data = session;
+	return nonseekable_open(inode, file);
+}
+
+static int vmcp_release(struct inode *inode, struct file *file)
+{
+	struct vmcp_session *session;
+
+	session = file->private_data;
+	file->private_data = NULL;
+	free_pages((unsigned long)session->response, get_order(session->bufsize));
+	kfree(session);
+	return 0;
+}
+
+static ssize_t
+vmcp_read(struct file *file, char __user *buff, size_t count, loff_t *ppos)
+{
+	ssize_t ret;
+	size_t size;
+	struct vmcp_session *session;
+
+	session = file->private_data;
+	if (mutex_lock_interruptible(&session->mutex))
+		return -ERESTARTSYS;
+	if (!session->response) {
+		mutex_unlock(&session->mutex);
+		return 0;
+	}
+	size = min_t(size_t, session->resp_size, session->bufsize);
+	ret = simple_read_from_buffer(buff, count, ppos,
+					session->response, size);
+
+	mutex_unlock(&session->mutex);
+
+	return ret;
+}
+
+static ssize_t
+vmcp_write(struct file *file, const char __user *buff, size_t count,
+	   loff_t *ppos)
+{
+	char *cmd;
+	struct vmcp_session *session;
+
+	if (count > 240)
+		return -EINVAL;
+	cmd = kmalloc(count + 1, GFP_KERNEL);
+	if (!cmd)
+		return -ENOMEM;
+	if (copy_from_user(cmd, buff, count)) {
+		kfree(cmd);
+		return -EFAULT;
+	}
+	cmd[count] = '\0';
+	session = file->private_data;
+	if (mutex_lock_interruptible(&session->mutex)) {
+		kfree(cmd);
+		return -ERESTARTSYS;
+	}
+	if (!session->response)
+		session->response = (char *)__get_free_pages(GFP_KERNEL
+						| __GFP_REPEAT | GFP_DMA,
+						get_order(session->bufsize));
+	if (!session->response) {
+		mutex_unlock(&session->mutex);
+		kfree(cmd);
+		return -ENOMEM;
+	}
+	debug_text_event(vmcp_debug, 1, cmd);
+	session->resp_size = cpcmd(cmd, session->response, session->bufsize,
+				   &session->resp_code);
+	mutex_unlock(&session->mutex);
+	kfree(cmd);
+	*ppos = 0;		/* reset the file pointer after a command */
+	return count;
+}
+
+
+/*
+ * These ioctls are available, as the semantics of the diagnose 8 call
+ * does not fit very well into a Linux call. Diagnose X'08' is described in
+ * CP Programming Services SC24-6084-00
+ *
+ * VMCP_GETCODE: gives the CP return code back to user space
+ * VMCP_SETBUF: sets the response buffer for the next write call. diagnose 8
+ * expects adjacent pages in real storage and to make matters worse, we
+ * dont know the size of the response. Therefore we default to PAGESIZE and
+ * let userspace to change the response size, if userspace expects a bigger
+ * response
+ */
+static long vmcp_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
+{
+	struct vmcp_session *session;
+	int __user *argp;
+	int temp;
+
+	session = file->private_data;
+	if (is_compat_task())
+		argp = compat_ptr(arg);
+	else
+		argp = (int __user *)arg;
+	if (mutex_lock_interruptible(&session->mutex))
+		return -ERESTARTSYS;
+	switch (cmd) {
+	case VMCP_GETCODE:
+		temp = session->resp_code;
+		mutex_unlock(&session->mutex);
+		return put_user(temp, argp);
+	case VMCP_SETBUF:
+		free_pages((unsigned long)session->response,
+				get_order(session->bufsize));
+		session->response=NULL;
+		temp = get_user(session->bufsize, argp);
+		if (get_order(session->bufsize) > 8) {
+			session->bufsize = PAGE_SIZE;
+			temp = -EINVAL;
+		}
+		mutex_unlock(&session->mutex);
+		return temp;
+	case VMCP_GETSIZE:
+		temp = session->resp_size;
+		mutex_unlock(&session->mutex);
+		return put_user(temp, argp);
+	default:
+		mutex_unlock(&session->mutex);
+		return -ENOIOCTLCMD;
+	}
+}
+
+static const struct file_operations vmcp_fops = {
+	.owner		= THIS_MODULE,
+	.open		= vmcp_open,
+	.release	= vmcp_release,
+	.read		= vmcp_read,
+	.write		= vmcp_write,
+	.unlocked_ioctl	= vmcp_ioctl,
+	.compat_ioctl	= vmcp_ioctl,
+	.llseek		= no_llseek,
+};
+
+static struct miscdevice vmcp_dev = {
+	.name	= "vmcp",
+	.minor	= MISC_DYNAMIC_MINOR,
+	.fops	= &vmcp_fops,
+};
+
+static int __init vmcp_init(void)
+{
+	int ret;
+
+	if (!MACHINE_IS_VM)
+		return 0;
+
+	vmcp_debug = debug_register("vmcp", 1, 1, 240);
+	if (!vmcp_debug)
+		return -ENOMEM;
+
+	ret = debug_register_view(vmcp_debug, &debug_hex_ascii_view);
+	if (ret) {
+		debug_unregister(vmcp_debug);
+		return ret;
+	}
+
+	ret = misc_register(&vmcp_dev);
+	if (ret)
+		debug_unregister(vmcp_debug);
+	return ret;
+}
+device_initcall(vmcp_init);

drivers/s390/char/vmcp.h

@@ -0,0 +1,30 @@
+/*
+ * Copyright IBM Corp. 2004, 2005
+ * Interface implementation for communication with the z/VM control program
+ * Version 1.0
+ * Author(s): Christian Borntraeger <cborntra@de.ibm.com>
+ *
+ *
+ * z/VMs CP offers the possibility to issue commands via the diagnose code 8
+ * this driver implements a character device that issues these commands and
+ * returns the answer of CP.
+ *
+ * The idea of this driver is based on cpint from Neale Ferguson
+ */
+
+#include <linux/ioctl.h>
+#include <linux/mutex.h>
+
+#define VMCP_GETCODE _IOR(0x10, 1, int)
+#define VMCP_SETBUF _IOW(0x10, 2, int)
+#define VMCP_GETSIZE _IOR(0x10, 3, int)
+
+struct vmcp_session {
+	unsigned int bufsize;
+	char *response;
+	int resp_size;
+	int resp_code;
+	/* As we use copy_from/to_user, which might     *
+	 * sleep and cannot use a spinlock              */
+	struct mutex mutex;
+};

drivers/s390/char/vmlogrdr.c

@@ -0,0 +1,907 @@
+/*
+ *	character device driver for reading z/VM system service records
+ *
+ *
+ *	Copyright IBM Corp. 2004, 2009
+ *	character device driver for reading z/VM system service records,
+ *	Version 1.0
+ *	Author(s): Xenia Tkatschow <xenia@us.ibm.com>
+ *		   Stefan Weinhuber <wein@de.ibm.com>
+ *
+ */
+
+#define KMSG_COMPONENT "vmlogrdr"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/slab.h>
+#include <linux/errno.h>
+#include <linux/types.h>
+#include <linux/interrupt.h>
+#include <linux/spinlock.h>
+#include <linux/atomic.h>
+#include <asm/uaccess.h>
+#include <asm/cpcmd.h>
+#include <asm/debug.h>
+#include <asm/ebcdic.h>
+#include <net/iucv/iucv.h>
+#include <linux/kmod.h>
+#include <linux/cdev.h>
+#include <linux/device.h>
+#include <linux/string.h>
+
+MODULE_AUTHOR
+	("(C) 2004 IBM Corporation by Xenia Tkatschow (xenia@us.ibm.com)\n"
+	 "                            Stefan Weinhuber (wein@de.ibm.com)");
+MODULE_DESCRIPTION ("Character device driver for reading z/VM "
+		    "system service records.");
+MODULE_LICENSE("GPL");
+
+
+/*
+ * The size of the buffer for iucv data transfer is one page,
+ * but in addition to the data we read from iucv we also
+ * place an integer and some characters into that buffer,
+ * so the maximum size for record data is a little less then
+ * one page.
+ */
+#define NET_BUFFER_SIZE	(PAGE_SIZE - sizeof(int) - sizeof(FENCE))
+
+/*
+ * The elements that are concurrently accessed by bottom halves are
+ * connection_established, iucv_path_severed, local_interrupt_buffer
+ * and receive_ready. The first three can be protected by
+ * priv_lock.  receive_ready is atomic, so it can be incremented and
+ * decremented without holding a lock.
+ * The variable dev_in_use needs to be protected by the lock, since
+ * it's a flag used by open to make sure that the device is opened only
+ * by one user at the same time.
+ */
+struct vmlogrdr_priv_t {
+	char system_service[8];
+	char internal_name[8];
+	char recording_name[8];
+	struct iucv_path *path;
+	int connection_established;
+	int iucv_path_severed;
+	struct iucv_message local_interrupt_buffer;
+	atomic_t receive_ready;
+	int minor_num;
+	char * buffer;
+	char * current_position;
+	int remaining;
+	ulong residual_length;
+	int buffer_free;
+	int dev_in_use; /* 1: already opened, 0: not opened*/
+	spinlock_t priv_lock;
+	struct device  *device;
+	struct device  *class_device;
+	int autorecording;
+	int autopurge;
+};
+
+
+/*
+ * File operation structure for vmlogrdr devices
+ */
+static int vmlogrdr_open(struct inode *, struct file *);
+static int vmlogrdr_release(struct inode *, struct file *);
+static ssize_t vmlogrdr_read (struct file *filp, char __user *data,
+			      size_t count, loff_t * ppos);
+
+static const struct file_operations vmlogrdr_fops = {
+	.owner   = THIS_MODULE,
+	.open    = vmlogrdr_open,
+	.release = vmlogrdr_release,
+	.read    = vmlogrdr_read,
+	.llseek  = no_llseek,
+};
+
+
+static void vmlogrdr_iucv_path_complete(struct iucv_path *, u8 ipuser[16]);
+static void vmlogrdr_iucv_path_severed(struct iucv_path *, u8 ipuser[16]);
+static void vmlogrdr_iucv_message_pending(struct iucv_path *,
+					  struct iucv_message *);
+
+
+static struct iucv_handler vmlogrdr_iucv_handler = {
+	.path_complete	 = vmlogrdr_iucv_path_complete,
+	.path_severed	 = vmlogrdr_iucv_path_severed,
+	.message_pending = vmlogrdr_iucv_message_pending,
+};
+
+
+static DECLARE_WAIT_QUEUE_HEAD(conn_wait_queue);
+static DECLARE_WAIT_QUEUE_HEAD(read_wait_queue);
+
+/*
+ * pointer to system service private structure
+ * minor number 0 --> logrec
+ * minor number 1 --> account
+ * minor number 2 --> symptom
+ */
+
+static struct vmlogrdr_priv_t sys_ser[] = {
+	{ .system_service = "*LOGREC ",
+	  .internal_name  = "logrec",
+	  .recording_name = "EREP",
+	  .minor_num      = 0,
+	  .buffer_free    = 1,
+	  .priv_lock	  = __SPIN_LOCK_UNLOCKED(sys_ser[0].priv_lock),
+	  .autorecording  = 1,
+	  .autopurge      = 1,
+	},
+	{ .system_service = "*ACCOUNT",
+	  .internal_name  = "account",
+	  .recording_name = "ACCOUNT",
+	  .minor_num      = 1,
+	  .buffer_free    = 1,
+	  .priv_lock	  = __SPIN_LOCK_UNLOCKED(sys_ser[1].priv_lock),
+	  .autorecording  = 1,
+	  .autopurge      = 1,
+	},
+	{ .system_service = "*SYMPTOM",
+	  .internal_name  = "symptom",
+	  .recording_name = "SYMPTOM",
+	  .minor_num      = 2,
+	  .buffer_free    = 1,
+	  .priv_lock	  = __SPIN_LOCK_UNLOCKED(sys_ser[2].priv_lock),
+	  .autorecording  = 1,
+	  .autopurge      = 1,
+	}
+};
+
+#define MAXMINOR  (sizeof(sys_ser)/sizeof(struct vmlogrdr_priv_t))
+
+static char FENCE[] = {"EOR"};
+static int vmlogrdr_major = 0;
+static struct cdev  *vmlogrdr_cdev = NULL;
+static int recording_class_AB;
+
+
+static void vmlogrdr_iucv_path_complete(struct iucv_path *path, u8 ipuser[16])
+{
+	struct vmlogrdr_priv_t * logptr = path->private;
+
+	spin_lock(&logptr->priv_lock);
+	logptr->connection_established = 1;
+	spin_unlock(&logptr->priv_lock);
+	wake_up(&conn_wait_queue);
+}
+
+
+static void vmlogrdr_iucv_path_severed(struct iucv_path *path, u8 ipuser[16])
+{
+	struct vmlogrdr_priv_t * logptr = path->private;
+	u8 reason = (u8) ipuser[8];
+
+	pr_err("vmlogrdr: connection severed with reason %i\n", reason);
+
+	iucv_path_sever(path, NULL);
+	kfree(path);
+	logptr->path = NULL;
+
+	spin_lock(&logptr->priv_lock);
+	logptr->connection_established = 0;
+	logptr->iucv_path_severed = 1;
+	spin_unlock(&logptr->priv_lock);
+
+	wake_up(&conn_wait_queue);
+	/* just in case we're sleeping waiting for a record */
+	wake_up_interruptible(&read_wait_queue);
+}
+
+
+static void vmlogrdr_iucv_message_pending(struct iucv_path *path,
+					  struct iucv_message *msg)
+{
+	struct vmlogrdr_priv_t * logptr = path->private;
+
+	/*
+	 * This function is the bottom half so it should be quick.
+	 * Copy the external interrupt data into our local eib and increment
+	 * the usage count
+	 */
+	spin_lock(&logptr->priv_lock);
+	memcpy(&logptr->local_interrupt_buffer, msg, sizeof(*msg));
+	atomic_inc(&logptr->receive_ready);
+	spin_unlock(&logptr->priv_lock);
+	wake_up_interruptible(&read_wait_queue);
+}
+
+
+static int vmlogrdr_get_recording_class_AB(void)
+{
+	static const char cp_command[] = "QUERY COMMAND RECORDING ";
+	char cp_response[80];
+	char *tail;
+	int len,i;
+
+	cpcmd(cp_command, cp_response, sizeof(cp_response), NULL);
+	len = strnlen(cp_response,sizeof(cp_response));
+	// now the parsing
+	tail=strnchr(cp_response,len,'=');
+	if (!tail)
+		return 0;
+	tail++;
+	if (!strncmp("ANY",tail,3))
+		return 1;
+	if (!strncmp("NONE",tail,4))
+		return 0;
+	/*
+	 * expect comma separated list of classes here, if one of them
+	 * is A or B return 1 otherwise 0
+	 */
+        for (i=tail-cp_response; i<len; i++)
+		if ( cp_response[i]=='A' || cp_response[i]=='B' )
+			return 1;
+	return 0;
+}
+
+
+static int vmlogrdr_recording(struct vmlogrdr_priv_t * logptr,
+			      int action, int purge)
+{
+
+	char cp_command[80];
+	char cp_response[160];
+	char *onoff, *qid_string;
+	int rc;
+
+	onoff = ((action == 1) ? "ON" : "OFF");
+	qid_string = ((recording_class_AB == 1) ? " QID * " : "");
+
+	/*
+	 * The recording commands needs to be called with option QID
+	 * for guests that have previlege classes A or B.
+	 * Purging has to be done as separate step, because recording
+	 * can't be switched on as long as records are on the queue.
+	 * Doing both at the same time doesn't work.
+	 */
+	if (purge && (action == 1)) {
+		memset(cp_command, 0x00, sizeof(cp_command));
+		memset(cp_response, 0x00, sizeof(cp_response));
+		snprintf(cp_command, sizeof(cp_command),
+			 "RECORDING %s PURGE %s",
+			 logptr->recording_name,
+			 qid_string);
+		cpcmd(cp_command, cp_response, sizeof(cp_response), NULL);
+	}
+
+	memset(cp_command, 0x00, sizeof(cp_command));
+	memset(cp_response, 0x00, sizeof(cp_response));
+	snprintf(cp_command, sizeof(cp_command), "RECORDING %s %s %s",
+		logptr->recording_name,
+		onoff,
+		qid_string);
+	cpcmd(cp_command, cp_response, sizeof(cp_response), NULL);
+	/* The recording command will usually answer with 'Command complete'
+	 * on success, but when the specific service was never connected
+	 * before then there might be an additional informational message
+	 * 'HCPCRC8072I Recording entry not found' before the
+	 * 'Command complete'. So I use strstr rather then the strncmp.
+	 */
+	if (strstr(cp_response,"Command complete"))
+		rc = 0;
+	else
+		rc = -EIO;
+	/*
+	 * If we turn recording off, we have to purge any remaining records
+	 * afterwards, as a large number of queued records may impact z/VM
+	 * performance.
+	 */
+	if (purge && (action == 0)) {
+		memset(cp_command, 0x00, sizeof(cp_command));
+		memset(cp_response, 0x00, sizeof(cp_response));
+		snprintf(cp_command, sizeof(cp_command),
+			 "RECORDING %s PURGE %s",
+			 logptr->recording_name,
+			 qid_string);
+		cpcmd(cp_command, cp_response, sizeof(cp_response), NULL);
+	}
+
+	return rc;
+}
+
+
+static int vmlogrdr_open (struct inode *inode, struct file *filp)
+{
+	int dev_num = 0;
+	struct vmlogrdr_priv_t * logptr = NULL;
+	int connect_rc = 0;
+	int ret;
+
+	dev_num = iminor(inode);
+	if (dev_num >= MAXMINOR)
+		return -ENODEV;
+	logptr = &sys_ser[dev_num];
+
+	/*
+	 * only allow for blocking reads to be open
+	 */
+	if (filp->f_flags & O_NONBLOCK)
+		return -EOPNOTSUPP;
+
+	/* Besure this device hasn't already been opened */
+	spin_lock_bh(&logptr->priv_lock);
+	if (logptr->dev_in_use)	{
+		spin_unlock_bh(&logptr->priv_lock);
+		return -EBUSY;
+	}
+	logptr->dev_in_use = 1;
+	logptr->connection_established = 0;
+	logptr->iucv_path_severed = 0;
+	atomic_set(&logptr->receive_ready, 0);
+	logptr->buffer_free = 1;
+	spin_unlock_bh(&logptr->priv_lock);
+
+	/* set the file options */
+	filp->private_data = logptr;
+
+	/* start recording for this service*/
+	if (logptr->autorecording) {
+		ret = vmlogrdr_recording(logptr,1,logptr->autopurge);
+		if (ret)
+			pr_warning("vmlogrdr: failed to start "
+				   "recording automatically\n");
+	}
+
+	/* create connection to the system service */
+	logptr->path = iucv_path_alloc(10, 0, GFP_KERNEL);
+	if (!logptr->path)
+		goto out_dev;
+	connect_rc = iucv_path_connect(logptr->path, &vmlogrdr_iucv_handler,
+				       logptr->system_service, NULL, NULL,
+				       logptr);
+	if (connect_rc) {
+		pr_err("vmlogrdr: iucv connection to %s "
+		       "failed with rc %i \n",
+		       logptr->system_service, connect_rc);
+		goto out_path;
+	}
+
+	/* We've issued the connect and now we must wait for a
+	 * ConnectionComplete or ConnectinSevered Interrupt
+	 * before we can continue to process.
+	 */
+	wait_event(conn_wait_queue, (logptr->connection_established)
+		   || (logptr->iucv_path_severed));
+	if (logptr->iucv_path_severed)
+		goto out_record;
+	nonseekable_open(inode, filp);
+	return 0;
+
+out_record:
+	if (logptr->autorecording)
+		vmlogrdr_recording(logptr,0,logptr->autopurge);
+out_path:
+	kfree(logptr->path);	/* kfree(NULL) is ok. */
+	logptr->path = NULL;
+out_dev:
+	logptr->dev_in_use = 0;
+	return -EIO;
+}
+
+
+static int vmlogrdr_release (struct inode *inode, struct file *filp)
+{
+	int ret;
+
+	struct vmlogrdr_priv_t * logptr = filp->private_data;
+
+	iucv_path_sever(logptr->path, NULL);
+	kfree(logptr->path);
+	logptr->path = NULL;
+	if (logptr->autorecording) {
+		ret = vmlogrdr_recording(logptr,0,logptr->autopurge);
+		if (ret)
+			pr_warning("vmlogrdr: failed to stop "
+				   "recording automatically\n");
+	}
+	logptr->dev_in_use = 0;
+
+	return 0;
+}
+
+
+static int vmlogrdr_receive_data(struct vmlogrdr_priv_t *priv)
+{
+	int rc, *temp;
+	/* we need to keep track of two data sizes here:
+	 * The number of bytes we need to receive from iucv and
+	 * the total number of bytes we actually write into the buffer.
+	 */
+	int user_data_count, iucv_data_count;
+	char * buffer;
+
+	if (atomic_read(&priv->receive_ready)) {
+		spin_lock_bh(&priv->priv_lock);
+		if (priv->residual_length){
+			/* receive second half of a record */
+			iucv_data_count = priv->residual_length;
+			user_data_count = 0;
+			buffer = priv->buffer;
+		} else {
+			/* receive a new record:
+			 * We need to return the total length of the record
+                         * + size of FENCE in the first 4 bytes of the buffer.
+		         */
+			iucv_data_count = priv->local_interrupt_buffer.length;
+			user_data_count = sizeof(int);
+			temp = (int*)priv->buffer;
+			*temp= iucv_data_count + sizeof(FENCE);
+			buffer = priv->buffer + sizeof(int);
+		}
+		/*
+		 * If the record is bigger than our buffer, we receive only
+		 * a part of it. We can get the rest later.
+		 */
+		if (iucv_data_count > NET_BUFFER_SIZE)
+			iucv_data_count = NET_BUFFER_SIZE;
+		rc = iucv_message_receive(priv->path,
+					  &priv->local_interrupt_buffer,
+					  0, buffer, iucv_data_count,
+					  &priv->residual_length);
+		spin_unlock_bh(&priv->priv_lock);
+		/* An rc of 5 indicates that the record was bigger than
+		 * the buffer, which is OK for us. A 9 indicates that the
+		 * record was purged befor we could receive it.
+		 */
+		if (rc == 5)
+			rc = 0;
+		if (rc == 9)
+			atomic_set(&priv->receive_ready, 0);
+	} else {
+		rc = 1;
+	}
+	if (!rc) {
+		priv->buffer_free = 0;
+ 		user_data_count += iucv_data_count;
+		priv->current_position = priv->buffer;
+		if (priv->residual_length == 0){
+			/* the whole record has been captured,
+			 * now add the fence */
+			atomic_dec(&priv->receive_ready);
+			buffer = priv->buffer + user_data_count;
+			memcpy(buffer, FENCE, sizeof(FENCE));
+			user_data_count += sizeof(FENCE);
+		}
+		priv->remaining = user_data_count;
+	}
+
+	return rc;
+}
+
+
+static ssize_t vmlogrdr_read(struct file *filp, char __user *data,
+			     size_t count, loff_t * ppos)
+{
+	int rc;
+	struct vmlogrdr_priv_t * priv = filp->private_data;
+
+	while (priv->buffer_free) {
+		rc = vmlogrdr_receive_data(priv);
+		if (rc) {
+			rc = wait_event_interruptible(read_wait_queue,
+					atomic_read(&priv->receive_ready));
+			if (rc)
+				return rc;
+		}
+	}
+	/* copy only up to end of record */
+	if (count > priv->remaining)
+		count = priv->remaining;
+
+	if (copy_to_user(data, priv->current_position, count))
+		return -EFAULT;
+
+	*ppos += count;
+	priv->current_position += count;
+	priv->remaining -= count;
+
+	/* if all data has been transferred, set buffer free */
+	if (priv->remaining == 0)
+		priv->buffer_free = 1;
+
+	return count;
+}
+
+static ssize_t vmlogrdr_autopurge_store(struct device * dev,
+					struct device_attribute *attr,
+					const char * buf, size_t count)
+{
+	struct vmlogrdr_priv_t *priv = dev_get_drvdata(dev);
+	ssize_t ret = count;
+
+	switch (buf[0]) {
+	case '0':
+		priv->autopurge=0;
+		break;
+	case '1':
+		priv->autopurge=1;
+		break;
+	default:
+		ret = -EINVAL;
+	}
+	return ret;
+}
+
+
+static ssize_t vmlogrdr_autopurge_show(struct device *dev,
+				       struct device_attribute *attr,
+				       char *buf)
+{
+	struct vmlogrdr_priv_t *priv = dev_get_drvdata(dev);
+	return sprintf(buf, "%u\n", priv->autopurge);
+}
+
+
+static DEVICE_ATTR(autopurge, 0644, vmlogrdr_autopurge_show,
+		   vmlogrdr_autopurge_store);
+
+
+static ssize_t vmlogrdr_purge_store(struct device * dev,
+				    struct device_attribute *attr,
+				    const char * buf, size_t count)
+{
+
+	char cp_command[80];
+	char cp_response[80];
+	struct vmlogrdr_priv_t *priv = dev_get_drvdata(dev);
+
+	if (buf[0] != '1')
+		return -EINVAL;
+
+	memset(cp_command, 0x00, sizeof(cp_command));
+	memset(cp_response, 0x00, sizeof(cp_response));
+
+        /*
+	 * The recording command needs to be called with option QID
+	 * for guests that have previlege classes A or B.
+	 * Other guests will not recognize the command and we have to
+	 * issue the same command without the QID parameter.
+	 */
+
+	if (recording_class_AB)
+		snprintf(cp_command, sizeof(cp_command),
+			 "RECORDING %s PURGE QID * ",
+			 priv->recording_name);
+	else
+		snprintf(cp_command, sizeof(cp_command),
+			 "RECORDING %s PURGE ",
+			 priv->recording_name);
+
+	cpcmd(cp_command, cp_response, sizeof(cp_response), NULL);
+
+	return count;
+}
+
+
+static DEVICE_ATTR(purge, 0200, NULL, vmlogrdr_purge_store);
+
+
+static ssize_t vmlogrdr_autorecording_store(struct device *dev,
+					    struct device_attribute *attr,
+					    const char *buf, size_t count)
+{
+	struct vmlogrdr_priv_t *priv = dev_get_drvdata(dev);
+	ssize_t ret = count;
+
+	switch (buf[0]) {
+	case '0':
+		priv->autorecording=0;
+		break;
+	case '1':
+		priv->autorecording=1;
+		break;
+	default:
+		ret = -EINVAL;
+	}
+	return ret;
+}
+
+
+static ssize_t vmlogrdr_autorecording_show(struct device *dev,
+					   struct device_attribute *attr,
+					   char *buf)
+{
+	struct vmlogrdr_priv_t *priv = dev_get_drvdata(dev);
+	return sprintf(buf, "%u\n", priv->autorecording);
+}
+
+
+static DEVICE_ATTR(autorecording, 0644, vmlogrdr_autorecording_show,
+		   vmlogrdr_autorecording_store);
+
+
+static ssize_t vmlogrdr_recording_store(struct device * dev,
+					struct device_attribute *attr,
+					const char * buf, size_t count)
+{
+	struct vmlogrdr_priv_t *priv = dev_get_drvdata(dev);
+	ssize_t ret;
+
+	switch (buf[0]) {
+	case '0':
+		ret = vmlogrdr_recording(priv,0,0);
+		break;
+	case '1':
+		ret = vmlogrdr_recording(priv,1,0);
+		break;
+	default:
+		ret = -EINVAL;
+	}
+	if (ret)
+		return ret;
+	else
+		return count;
+
+}
+
+
+static DEVICE_ATTR(recording, 0200, NULL, vmlogrdr_recording_store);
+
+
+static ssize_t vmlogrdr_recording_status_show(struct device_driver *driver,
+					      char *buf)
+{
+
+	static const char cp_command[] = "QUERY RECORDING ";
+	int len;
+
+	cpcmd(cp_command, buf, 4096, NULL);
+	len = strlen(buf);
+	return len;
+}
+static DRIVER_ATTR(recording_status, 0444, vmlogrdr_recording_status_show,
+		   NULL);
+static struct attribute *vmlogrdr_drv_attrs[] = {
+	&driver_attr_recording_status.attr,
+	NULL,
+};
+static struct attribute_group vmlogrdr_drv_attr_group = {
+	.attrs = vmlogrdr_drv_attrs,
+};
+static const struct attribute_group *vmlogrdr_drv_attr_groups[] = {
+	&vmlogrdr_drv_attr_group,
+	NULL,
+};
+
+static struct attribute *vmlogrdr_attrs[] = {
+	&dev_attr_autopurge.attr,
+	&dev_attr_purge.attr,
+	&dev_attr_autorecording.attr,
+	&dev_attr_recording.attr,
+	NULL,
+};
+static struct attribute_group vmlogrdr_attr_group = {
+	.attrs = vmlogrdr_attrs,
+};
+static const struct attribute_group *vmlogrdr_attr_groups[] = {
+	&vmlogrdr_attr_group,
+	NULL,
+};
+
+static int vmlogrdr_pm_prepare(struct device *dev)
+{
+	int rc;
+	struct vmlogrdr_priv_t *priv = dev_get_drvdata(dev);
+
+	rc = 0;
+	if (priv) {
+		spin_lock_bh(&priv->priv_lock);
+		if (priv->dev_in_use)
+			rc = -EBUSY;
+		spin_unlock_bh(&priv->priv_lock);
+	}
+	if (rc)
+		pr_err("vmlogrdr: device %s is busy. Refuse to suspend.\n",
+		       dev_name(dev));
+	return rc;
+}
+
+
+static const struct dev_pm_ops vmlogrdr_pm_ops = {
+	.prepare = vmlogrdr_pm_prepare,
+};
+
+static struct class *vmlogrdr_class;
+static struct device_driver vmlogrdr_driver = {
+	.name = "vmlogrdr",
+	.bus  = &iucv_bus,
+	.pm = &vmlogrdr_pm_ops,
+	.groups = vmlogrdr_drv_attr_groups,
+};
+
+static int vmlogrdr_register_driver(void)
+{
+	int ret;
+
+	/* Register with iucv driver */
+	ret = iucv_register(&vmlogrdr_iucv_handler, 1);
+	if (ret)
+		goto out;
+
+	ret = driver_register(&vmlogrdr_driver);
+	if (ret)
+		goto out_iucv;
+
+	vmlogrdr_class = class_create(THIS_MODULE, "vmlogrdr");
+	if (IS_ERR(vmlogrdr_class)) {
+		ret = PTR_ERR(vmlogrdr_class);
+		vmlogrdr_class = NULL;
+		goto out_driver;
+	}
+	return 0;
+
+out_driver:
+	driver_unregister(&vmlogrdr_driver);
+out_iucv:
+	iucv_unregister(&vmlogrdr_iucv_handler, 1);
+out:
+	return ret;
+}
+
+
+static void vmlogrdr_unregister_driver(void)
+{
+	class_destroy(vmlogrdr_class);
+	vmlogrdr_class = NULL;
+	driver_unregister(&vmlogrdr_driver);
+	iucv_unregister(&vmlogrdr_iucv_handler, 1);
+}
+
+
+static int vmlogrdr_register_device(struct vmlogrdr_priv_t *priv)
+{
+	struct device *dev;
+	int ret;
+
+	dev = kzalloc(sizeof(struct device), GFP_KERNEL);
+	if (dev) {
+		dev_set_name(dev, "%s", priv->internal_name);
+		dev->bus = &iucv_bus;
+		dev->parent = iucv_root;
+		dev->driver = &vmlogrdr_driver;
+		dev->groups = vmlogrdr_attr_groups;
+		dev_set_drvdata(dev, priv);
+		/*
+		 * The release function could be called after the
+		 * module has been unloaded. It's _only_ task is to
+		 * free the struct. Therefore, we specify kfree()
+		 * directly here. (Probably a little bit obfuscating
+		 * but legitime ...).
+		 */
+		dev->release = (void (*)(struct device *))kfree;
+	} else
+		return -ENOMEM;
+	ret = device_register(dev);
+	if (ret) {
+		put_device(dev);
+		return ret;
+	}
+
+	priv->class_device = device_create(vmlogrdr_class, dev,
+					   MKDEV(vmlogrdr_major,
+						 priv->minor_num),
+					   priv, "%s", dev_name(dev));
+	if (IS_ERR(priv->class_device)) {
+		ret = PTR_ERR(priv->class_device);
+		priv->class_device=NULL;
+		device_unregister(dev);
+		return ret;
+	}
+	priv->device = dev;
+	return 0;
+}
+
+
+static int vmlogrdr_unregister_device(struct vmlogrdr_priv_t *priv)
+{
+	device_destroy(vmlogrdr_class, MKDEV(vmlogrdr_major, priv->minor_num));
+	if (priv->device != NULL) {
+		device_unregister(priv->device);
+		priv->device=NULL;
+	}
+	return 0;
+}
+
+
+static int vmlogrdr_register_cdev(dev_t dev)
+{
+	int rc = 0;
+	vmlogrdr_cdev = cdev_alloc();
+	if (!vmlogrdr_cdev) {
+		return -ENOMEM;
+	}
+	vmlogrdr_cdev->owner = THIS_MODULE;
+	vmlogrdr_cdev->ops = &vmlogrdr_fops;
+	vmlogrdr_cdev->dev = dev;
+	rc = cdev_add(vmlogrdr_cdev, vmlogrdr_cdev->dev, MAXMINOR);
+	if (!rc)
+		return 0;
+
+	// cleanup: cdev is not fully registered, no cdev_del here!
+	kobject_put(&vmlogrdr_cdev->kobj);
+	vmlogrdr_cdev=NULL;
+	return rc;
+}
+
+
+static void vmlogrdr_cleanup(void)
+{
+        int i;
+
+	if (vmlogrdr_cdev) {
+		cdev_del(vmlogrdr_cdev);
+		vmlogrdr_cdev=NULL;
+	}
+	for (i=0; i < MAXMINOR; ++i ) {
+		vmlogrdr_unregister_device(&sys_ser[i]);
+		free_page((unsigned long)sys_ser[i].buffer);
+	}
+	vmlogrdr_unregister_driver();
+	if (vmlogrdr_major) {
+		unregister_chrdev_region(MKDEV(vmlogrdr_major, 0), MAXMINOR);
+		vmlogrdr_major=0;
+	}
+}
+
+
+static int __init vmlogrdr_init(void)
+{
+	int rc;
+	int i;
+	dev_t dev;
+
+	if (! MACHINE_IS_VM) {
+		pr_err("not running under VM, driver not loaded.\n");
+		return -ENODEV;
+	}
+
+        recording_class_AB = vmlogrdr_get_recording_class_AB();
+
+	rc = alloc_chrdev_region(&dev, 0, MAXMINOR, "vmlogrdr");
+	if (rc)
+		return rc;
+	vmlogrdr_major = MAJOR(dev);
+
+	rc=vmlogrdr_register_driver();
+	if (rc)
+		goto cleanup;
+
+	for (i=0; i < MAXMINOR; ++i ) {
+		sys_ser[i].buffer = (char *) get_zeroed_page(GFP_KERNEL);
+		if (!sys_ser[i].buffer) {
+			rc = -ENOMEM;
+			break;
+		}
+		sys_ser[i].current_position = sys_ser[i].buffer;
+		rc=vmlogrdr_register_device(&sys_ser[i]);
+		if (rc)
+			break;
+	}
+	if (rc)
+		goto cleanup;
+
+	rc = vmlogrdr_register_cdev(dev);
+	if (rc)
+		goto cleanup;
+	return 0;
+
+cleanup:
+	vmlogrdr_cleanup();
+	return rc;
+}
+
+
+static void __exit vmlogrdr_exit(void)
+{
+	vmlogrdr_cleanup();
+	return;
+}
+
+
+module_init(vmlogrdr_init);
+module_exit(vmlogrdr_exit);

drivers/s390/char/vmur.h

@@ -0,0 +1,110 @@
+/*
+ * Linux driver for System z and s390 unit record devices
+ * (z/VM virtual punch, reader, printer)
+ *
+ * Copyright IBM Corp. 2001, 2007
+ * Authors: Malcolm Beattie <beattiem@uk.ibm.com>
+ *	    Michael Holzheu <holzheu@de.ibm.com>
+ *	    Frank Munzert <munzert@de.ibm.com>
+ */
+
+#ifndef _VMUR_H_
+#define _VMUR_H_
+
+#define DEV_CLASS_UR_I 0x20 /* diag210 unit record input device class */
+#define DEV_CLASS_UR_O 0x10 /* diag210 unit record output device class */
+/*
+ * we only support z/VM's default unit record devices:
+ * both in SPOOL directory control statement and in CP DEFINE statement
+ *	RDR defaults to 2540 reader
+ *	PUN defaults to 2540 punch
+ *	PRT defaults to 1403 printer
+ */
+#define READER_PUNCH_DEVTYPE	0x2540
+#define PRINTER_DEVTYPE		0x1403
+
+/* z/VM spool file control block SFBLOK */
+struct file_control_block {
+	char reserved_1[8];
+	char user_owner[8];
+	char user_orig[8];
+	__s32 data_recs;
+	__s16 rec_len;
+	__s16 file_num;
+	__u8  file_stat;
+	__u8  dev_type;
+	char  reserved_2[6];
+	char  file_name[12];
+	char  file_type[12];
+	char  create_date[8];
+	char  create_time[8];
+	char  reserved_3[6];
+	__u8  file_class;
+	__u8  sfb_lok;
+	__u64 distr_code;
+	__u32 reserved_4;
+	__u8  current_starting_copy_number;
+	__u8  sfblock_cntrl_flags;
+	__u8  reserved_5;
+	__u8  more_status_flags;
+	char  rest[200];
+} __attribute__ ((packed));
+
+#define FLG_SYSTEM_HOLD	0x04
+#define FLG_CP_DUMP	0x10
+#define FLG_USER_HOLD	0x20
+#define FLG_IN_USE	0x80
+
+/*
+ * A struct urdev is created for each ur device that is made available
+ * via the ccw_device driver model.
+ */
+struct urdev {
+	struct ccw_device *cdev;	/* Backpointer to ccw device */
+	struct mutex io_mutex;		/* Serialises device IO */
+	struct completion *io_done;	/* do_ur_io waits; irq completes */
+	struct device *device;
+	struct cdev *char_device;
+	struct ccw_dev_id dev_id;	/* device id */
+	size_t reclen;			/* Record length for *write* CCWs */
+	int class;			/* VM device class */
+	int io_request_rc;		/* return code from I/O request */
+	atomic_t ref_count;		/* reference counter */
+	wait_queue_head_t wait;		/* wait queue to serialize open */
+	int open_flag;			/* "urdev is open" flag */
+	spinlock_t open_lock;		/* serialize critical sections */
+};
+
+/*
+ * A struct urfile is allocated at open() time for each device and
+ * freed on release().
+ */
+struct urfile {
+	struct urdev *urd;
+	unsigned int flags;
+	size_t dev_reclen;
+	__u16 file_reclen;
+};
+
+/*
+ * Device major/minor definitions.
+ */
+
+#define UR_MAJOR 0	/* get dynamic major */
+/*
+ * We map minor numbers directly to device numbers (0-FFFF) for simplicity.
+ * This avoids having to allocate (and manage) slot numbers.
+ */
+#define NUM_MINORS 65536
+
+/* Limiting each I/O to 511 records limits chan prog to 4KB (511 r/w + 1 NOP) */
+#define MAX_RECS_PER_IO		511
+#define WRITE_CCW_CMD		0x01
+
+#define TRACE(x...) debug_sprintf_event(vmur_dbf, 1, x)
+#define CCWDEV_CU_DI(cutype, di) \
+		CCW_DEVICE(cutype, 0x00), .driver_info = (di)
+
+#define FILE_RECLEN_OFFSET	4064 /* reclen offset in spool data block */
+
+#endif /* _VMUR_H_ */

drivers/s390/char/zcore.c

@@ -0,0 +1,782 @@
+/*
+ * zcore module to export memory content and register sets for creating system
+ * dumps on SCSI disks (zfcpdump). The "zcore/mem" debugfs file shows the same
+ * dump format as s390 standalone dumps.
+ *
+ * For more information please refer to Documentation/s390/zfcpdump.txt
+ *
+ * Copyright IBM Corp. 2003, 2008
+ * Author(s): Michael Holzheu
+ */
+
+#define KMSG_COMPONENT "zdump"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/init.h>
+#include <linux/slab.h>
+#include <linux/miscdevice.h>
+#include <linux/debugfs.h>
+#include <linux/module.h>
+#include <linux/memblock.h>
+
+#include <asm/asm-offsets.h>
+#include <asm/ipl.h>
+#include <asm/sclp.h>
+#include <asm/setup.h>
+#include <asm/uaccess.h>
+#include <asm/debug.h>
+#include <asm/processor.h>
+#include <asm/irqflags.h>
+#include <asm/checksum.h>
+#include <asm/switch_to.h>
+#include "sclp.h"
+
+#define TRACE(x...) debug_sprintf_event(zcore_dbf, 1, x)
+
+#define TO_USER		1
+#define TO_KERNEL	0
+#define CHUNK_INFO_SIZE	34 /* 2 16-byte char, each followed by blank */
+
+enum arch_id {
+	ARCH_S390	= 0,
+	ARCH_S390X	= 1,
+};
+
+/* dump system info */
+
+struct sys_info {
+	enum arch_id	 arch;
+	unsigned long	 sa_base;
+	u32		 sa_size;
+	int		 cpu_map[NR_CPUS];
+	unsigned long	 mem_size;
+	struct save_area lc_mask;
+};
+
+struct ipib_info {
+	unsigned long	ipib;
+	u32		checksum;
+}  __attribute__((packed));
+
+static struct sys_info sys_info;
+static struct debug_info *zcore_dbf;
+static int hsa_available;
+static struct dentry *zcore_dir;
+static struct dentry *zcore_file;
+static struct dentry *zcore_memmap_file;
+static struct dentry *zcore_reipl_file;
+static struct dentry *zcore_hsa_file;
+static struct ipl_parameter_block *ipl_block;
+
+/*
+ * Copy memory from HSA to kernel or user memory (not reentrant):
+ *
+ * @dest:  Kernel or user buffer where memory should be copied to
+ * @src:   Start address within HSA where data should be copied
+ * @count: Size of buffer, which should be copied
+ * @mode:  Either TO_KERNEL or TO_USER
+ */
+int memcpy_hsa(void *dest, unsigned long src, size_t count, int mode)
+{
+	int offs, blk_num;
+	static char buf[PAGE_SIZE] __attribute__((__aligned__(PAGE_SIZE)));
+
+	if (!hsa_available)
+		return -ENODATA;
+	if (count == 0)
+		return 0;
+
+	/* copy first block */
+	offs = 0;
+	if ((src % PAGE_SIZE) != 0) {
+		blk_num = src / PAGE_SIZE + 2;
+		if (sclp_sdias_copy(buf, blk_num, 1)) {
+			TRACE("sclp_sdias_copy() failed\n");
+			return -EIO;
+		}
+		offs = min((PAGE_SIZE - (src % PAGE_SIZE)), count);
+		if (mode == TO_USER) {
+			if (copy_to_user((__force __user void*) dest,
+					 buf + (src % PAGE_SIZE), offs))
+				return -EFAULT;
+		} else
+			memcpy(dest, buf + (src % PAGE_SIZE), offs);
+	}
+	if (offs == count)
+		goto out;
+
+	/* copy middle */
+	for (; (offs + PAGE_SIZE) <= count; offs += PAGE_SIZE) {
+		blk_num = (src + offs) / PAGE_SIZE + 2;
+		if (sclp_sdias_copy(buf, blk_num, 1)) {
+			TRACE("sclp_sdias_copy() failed\n");
+			return -EIO;
+		}
+		if (mode == TO_USER) {
+			if (copy_to_user((__force __user void*) dest + offs,
+					 buf, PAGE_SIZE))
+				return -EFAULT;
+		} else
+			memcpy(dest + offs, buf, PAGE_SIZE);
+	}
+	if (offs == count)
+		goto out;
+
+	/* copy last block */
+	blk_num = (src + offs) / PAGE_SIZE + 2;
+	if (sclp_sdias_copy(buf, blk_num, 1)) {
+		TRACE("sclp_sdias_copy() failed\n");
+		return -EIO;
+	}
+	if (mode == TO_USER) {
+		if (copy_to_user((__force __user void*) dest + offs, buf,
+				 count - offs))
+			return -EFAULT;
+	} else
+		memcpy(dest + offs, buf, count - offs);
+out:
+	return 0;
+}
+
+static int memcpy_hsa_user(void __user *dest, unsigned long src, size_t count)
+{
+	return memcpy_hsa((void __force *) dest, src, count, TO_USER);
+}
+
+static int memcpy_hsa_kernel(void *dest, unsigned long src, size_t count)
+{
+	return memcpy_hsa(dest, src, count, TO_KERNEL);
+}
+
+static int __init init_cpu_info(enum arch_id arch)
+{
+	struct save_area_ext *sa_ext;
+
+	/* get info for boot cpu from lowcore, stored in the HSA */
+
+	sa_ext = dump_save_area_create(0);
+	if (!sa_ext)
+		return -ENOMEM;
+	if (memcpy_hsa_kernel(&sa_ext->sa, sys_info.sa_base,
+			      sys_info.sa_size) < 0) {
+		TRACE("could not copy from HSA\n");
+		kfree(sa_ext);
+		return -EIO;
+	}
+	if (MACHINE_HAS_VX)
+		save_vx_regs_safe(sa_ext->vx_regs);
+	return 0;
+}
+
+static DEFINE_MUTEX(zcore_mutex);
+
+#define DUMP_VERSION	0x5
+#define DUMP_MAGIC	0xa8190173618f23fdULL
+#define DUMP_ARCH_S390X	2
+#define DUMP_ARCH_S390	1
+#define HEADER_SIZE	4096
+
+/* dump header dumped according to s390 crash dump format */
+
+struct zcore_header {
+	u64 magic;
+	u32 version;
+	u32 header_size;
+	u32 dump_level;
+	u32 page_size;
+	u64 mem_size;
+	u64 mem_start;
+	u64 mem_end;
+	u32 num_pages;
+	u32 pad1;
+	u64 tod;
+	struct cpuid cpu_id;
+	u32 arch_id;
+	u32 volnr;
+	u32 build_arch;
+	u64 rmem_size;
+	u8 mvdump;
+	u16 cpu_cnt;
+	u16 real_cpu_cnt;
+	u8 end_pad1[0x200-0x061];
+	u64 mvdump_sign;
+	u64 mvdump_zipl_time;
+	u8 end_pad2[0x800-0x210];
+	u32 lc_vec[512];
+} __attribute__((packed,__aligned__(16)));
+
+static struct zcore_header zcore_header = {
+	.magic		= DUMP_MAGIC,
+	.version	= DUMP_VERSION,
+	.header_size	= 4096,
+	.dump_level	= 0,
+	.page_size	= PAGE_SIZE,
+	.mem_start	= 0,
+#ifdef CONFIG_64BIT
+	.build_arch	= DUMP_ARCH_S390X,
+#else
+	.build_arch	= DUMP_ARCH_S390,
+#endif
+};
+
+/*
+ * Copy lowcore info to buffer. Use map in order to copy only register parts.
+ *
+ * @buf:    User buffer
+ * @sa:     Pointer to save area
+ * @sa_off: Offset in save area to copy
+ * @len:    Number of bytes to copy
+ */
+static int copy_lc(void __user *buf, void *sa, int sa_off, int len)
+{
+	int i;
+	char *lc_mask = (char*)&sys_info.lc_mask;
+
+	for (i = 0; i < len; i++) {
+		if (!lc_mask[i + sa_off])
+			continue;
+		if (copy_to_user(buf + i, sa + sa_off + i, 1))
+			return -EFAULT;
+	}
+	return 0;
+}
+
+/*
+ * Copy lowcores info to memory, if necessary
+ *
+ * @buf:   User buffer
+ * @addr:  Start address of buffer in dump memory
+ * @count: Size of buffer
+ */
+static int zcore_add_lc(char __user *buf, unsigned long start, size_t count)
+{
+	unsigned long end;
+	int i;
+
+	if (count == 0)
+		return 0;
+
+	end = start + count;
+	for (i = 0; i < dump_save_areas.count; i++) {
+		unsigned long cp_start, cp_end; /* copy range */
+		unsigned long sa_start, sa_end; /* save area range */
+		unsigned long prefix;
+		unsigned long sa_off, len, buf_off;
+		struct save_area *save_area = &dump_save_areas.areas[i]->sa;
+
+		prefix = save_area->pref_reg;
+		sa_start = prefix + sys_info.sa_base;
+		sa_end = prefix + sys_info.sa_base + sys_info.sa_size;
+
+		if ((end < sa_start) || (start > sa_end))
+			continue;
+		cp_start = max(start, sa_start);
+		cp_end = min(end, sa_end);
+
+		buf_off = cp_start - start;
+		sa_off = cp_start - sa_start;
+		len = cp_end - cp_start;
+
+		TRACE("copy_lc for: %lx\n", start);
+		if (copy_lc(buf + buf_off, save_area, sa_off, len))
+			return -EFAULT;
+	}
+	return 0;
+}
+
+/*
+ * Release the HSA
+ */
+static void release_hsa(void)
+{
+	diag308(DIAG308_REL_HSA, NULL);
+	hsa_available = 0;
+}
+
+/*
+ * Read routine for zcore character device
+ * First 4K are dump header
+ * Next 32MB are HSA Memory
+ * Rest is read from absolute Memory
+ */
+static ssize_t zcore_read(struct file *file, char __user *buf, size_t count,
+			  loff_t *ppos)
+{
+	unsigned long mem_start; /* Start address in memory */
+	size_t mem_offs;	 /* Offset in dump memory */
+	size_t hdr_count;	 /* Size of header part of output buffer */
+	size_t size;
+	int rc;
+
+	mutex_lock(&zcore_mutex);
+
+	if (*ppos > (sys_info.mem_size + HEADER_SIZE)) {
+		rc = -EINVAL;
+		goto fail;
+	}
+
+	count = min(count, (size_t) (sys_info.mem_size + HEADER_SIZE - *ppos));
+
+	/* Copy dump header */
+	if (*ppos < HEADER_SIZE) {
+		size = min(count, (size_t) (HEADER_SIZE - *ppos));
+		if (copy_to_user(buf, &zcore_header + *ppos, size)) {
+			rc = -EFAULT;
+			goto fail;
+		}
+		hdr_count = size;
+		mem_start = 0;
+	} else {
+		hdr_count = 0;
+		mem_start = *ppos - HEADER_SIZE;
+	}
+
+	mem_offs = 0;
+
+	/* Copy from HSA data */
+	if (*ppos < sclp_get_hsa_size() + HEADER_SIZE) {
+		size = min((count - hdr_count),
+			   (size_t) (sclp_get_hsa_size() - mem_start));
+		rc = memcpy_hsa_user(buf + hdr_count, mem_start, size);
+		if (rc)
+			goto fail;
+
+		mem_offs += size;
+	}
+
+	/* Copy from real mem */
+	size = count - mem_offs - hdr_count;
+	rc = copy_to_user_real(buf + hdr_count + mem_offs,
+			       (void *) mem_start + mem_offs, size);
+	if (rc)
+		goto fail;
+
+	/*
+	 * Since s390 dump analysis tools like lcrash or crash
+	 * expect register sets in the prefix pages of the cpus,
+	 * we copy them into the read buffer, if necessary.
+	 * buf + hdr_count: Start of memory part of output buffer
+	 * mem_start: Start memory address to copy from
+	 * count - hdr_count: Size of memory area to copy
+	 */
+	if (zcore_add_lc(buf + hdr_count, mem_start, count - hdr_count)) {
+		rc = -EFAULT;
+		goto fail;
+	}
+	*ppos += count;
+fail:
+	mutex_unlock(&zcore_mutex);
+	return (rc < 0) ? rc : count;
+}
+
+static int zcore_open(struct inode *inode, struct file *filp)
+{
+	if (!hsa_available)
+		return -ENODATA;
+	else
+		return capable(CAP_SYS_RAWIO) ? 0 : -EPERM;
+}
+
+static int zcore_release(struct inode *inode, struct file *filep)
+{
+	if (hsa_available)
+		release_hsa();
+	return 0;
+}
+
+static loff_t zcore_lseek(struct file *file, loff_t offset, int orig)
+{
+	loff_t rc;
+
+	mutex_lock(&zcore_mutex);
+	switch (orig) {
+	case 0:
+		file->f_pos = offset;
+		rc = file->f_pos;
+		break;
+	case 1:
+		file->f_pos += offset;
+		rc = file->f_pos;
+		break;
+	default:
+		rc = -EINVAL;
+	}
+	mutex_unlock(&zcore_mutex);
+	return rc;
+}
+
+static const struct file_operations zcore_fops = {
+	.owner		= THIS_MODULE,
+	.llseek		= zcore_lseek,
+	.read		= zcore_read,
+	.open		= zcore_open,
+	.release	= zcore_release,
+};
+
+static ssize_t zcore_memmap_read(struct file *filp, char __user *buf,
+				 size_t count, loff_t *ppos)
+{
+	return simple_read_from_buffer(buf, count, ppos, filp->private_data,
+				       memblock.memory.cnt * CHUNK_INFO_SIZE);
+}
+
+static int zcore_memmap_open(struct inode *inode, struct file *filp)
+{
+	struct memblock_region *reg;
+	char *buf;
+	int i = 0;
+
+	buf = kzalloc(memblock.memory.cnt * CHUNK_INFO_SIZE, GFP_KERNEL);
+	if (!buf) {
+		return -ENOMEM;
+	}
+	for_each_memblock(memory, reg) {
+		sprintf(buf + (i++ * CHUNK_INFO_SIZE), "%016llx %016llx ",
+			(unsigned long long) reg->base,
+			(unsigned long long) reg->size);
+	}
+	filp->private_data = buf;
+	return nonseekable_open(inode, filp);
+}
+
+static int zcore_memmap_release(struct inode *inode, struct file *filp)
+{
+	kfree(filp->private_data);
+	return 0;
+}
+
+static const struct file_operations zcore_memmap_fops = {
+	.owner		= THIS_MODULE,
+	.read		= zcore_memmap_read,
+	.open		= zcore_memmap_open,
+	.release	= zcore_memmap_release,
+	.llseek		= no_llseek,
+};
+
+static ssize_t zcore_reipl_write(struct file *filp, const char __user *buf,
+				 size_t count, loff_t *ppos)
+{
+	if (ipl_block) {
+		diag308(DIAG308_SET, ipl_block);
+		diag308(DIAG308_IPL, NULL);
+	}
+	return count;
+}
+
+static int zcore_reipl_open(struct inode *inode, struct file *filp)
+{
+	return nonseekable_open(inode, filp);
+}
+
+static int zcore_reipl_release(struct inode *inode, struct file *filp)
+{
+	return 0;
+}
+
+static const struct file_operations zcore_reipl_fops = {
+	.owner		= THIS_MODULE,
+	.write		= zcore_reipl_write,
+	.open		= zcore_reipl_open,
+	.release	= zcore_reipl_release,
+	.llseek		= no_llseek,
+};
+
+static ssize_t zcore_hsa_read(struct file *filp, char __user *buf,
+			      size_t count, loff_t *ppos)
+{
+	static char str[18];
+
+	if (hsa_available)
+		snprintf(str, sizeof(str), "%lx\n", sclp_get_hsa_size());
+	else
+		snprintf(str, sizeof(str), "0\n");
+	return simple_read_from_buffer(buf, count, ppos, str, strlen(str));
+}
+
+static ssize_t zcore_hsa_write(struct file *filp, const char __user *buf,
+			       size_t count, loff_t *ppos)
+{
+	char value;
+
+	if (*ppos != 0)
+		return -EPIPE;
+	if (copy_from_user(&value, buf, 1))
+		return -EFAULT;
+	if (value != '0')
+		return -EINVAL;
+	release_hsa();
+	return count;
+}
+
+static const struct file_operations zcore_hsa_fops = {
+	.owner		= THIS_MODULE,
+	.write		= zcore_hsa_write,
+	.read		= zcore_hsa_read,
+	.open		= nonseekable_open,
+	.llseek		= no_llseek,
+};
+
+#ifdef CONFIG_32BIT
+
+static void __init set_lc_mask(struct save_area *map)
+{
+	memset(&map->ext_save, 0xff, sizeof(map->ext_save));
+	memset(&map->timer, 0xff, sizeof(map->timer));
+	memset(&map->clk_cmp, 0xff, sizeof(map->clk_cmp));
+	memset(&map->psw, 0xff, sizeof(map->psw));
+	memset(&map->pref_reg, 0xff, sizeof(map->pref_reg));
+	memset(&map->acc_regs, 0xff, sizeof(map->acc_regs));
+	memset(&map->fp_regs, 0xff, sizeof(map->fp_regs));
+	memset(&map->gp_regs, 0xff, sizeof(map->gp_regs));
+	memset(&map->ctrl_regs, 0xff, sizeof(map->ctrl_regs));
+}
+
+#else /* CONFIG_32BIT */
+
+static void __init set_lc_mask(struct save_area *map)
+{
+	memset(&map->fp_regs, 0xff, sizeof(map->fp_regs));
+	memset(&map->gp_regs, 0xff, sizeof(map->gp_regs));
+	memset(&map->psw, 0xff, sizeof(map->psw));
+	memset(&map->pref_reg, 0xff, sizeof(map->pref_reg));
+	memset(&map->fp_ctrl_reg, 0xff, sizeof(map->fp_ctrl_reg));
+	memset(&map->tod_reg, 0xff, sizeof(map->tod_reg));
+	memset(&map->timer, 0xff, sizeof(map->timer));
+	memset(&map->clk_cmp, 0xff, sizeof(map->clk_cmp));
+	memset(&map->acc_regs, 0xff, sizeof(map->acc_regs));
+	memset(&map->ctrl_regs, 0xff, sizeof(map->ctrl_regs));
+}
+
+#endif /* CONFIG_32BIT */
+
+/*
+ * Initialize dump globals for a given architecture
+ */
+static int __init sys_info_init(enum arch_id arch, unsigned long mem_end)
+{
+	int rc;
+
+	switch (arch) {
+	case ARCH_S390X:
+		pr_alert("DETECTED 'S390X (64 bit) OS'\n");
+		break;
+	case ARCH_S390:
+		pr_alert("DETECTED 'S390 (32 bit) OS'\n");
+		break;
+	default:
+		pr_alert("0x%x is an unknown architecture.\n",arch);
+		return -EINVAL;
+	}
+	sys_info.sa_base = SAVE_AREA_BASE;
+	sys_info.sa_size = sizeof(struct save_area);
+	sys_info.arch = arch;
+	set_lc_mask(&sys_info.lc_mask);
+	rc = init_cpu_info(arch);
+	if (rc)
+		return rc;
+	sys_info.mem_size = mem_end;
+
+	return 0;
+}
+
+static int __init check_sdias(void)
+{
+	if (!sclp_get_hsa_size()) {
+		TRACE("Could not determine HSA size\n");
+		return -ENODEV;
+	}
+	return 0;
+}
+
+static int __init get_mem_info(unsigned long *mem, unsigned long *end)
+{
+	struct memblock_region *reg;
+
+	for_each_memblock(memory, reg) {
+		*mem += reg->size;
+		*end = max_t(unsigned long, *end, reg->base + reg->size);
+	}
+	return 0;
+}
+
+static void __init zcore_header_init(int arch, struct zcore_header *hdr,
+				     unsigned long mem_size)
+{
+	u32 prefix;
+	int i;
+
+	if (arch == ARCH_S390X)
+		hdr->arch_id = DUMP_ARCH_S390X;
+	else
+		hdr->arch_id = DUMP_ARCH_S390;
+	hdr->mem_size = mem_size;
+	hdr->rmem_size = mem_size;
+	hdr->mem_end = sys_info.mem_size;
+	hdr->num_pages = mem_size / PAGE_SIZE;
+	hdr->tod = get_tod_clock();
+	get_cpu_id(&hdr->cpu_id);
+	for (i = 0; i < dump_save_areas.count; i++) {
+		prefix = dump_save_areas.areas[i]->sa.pref_reg;
+		hdr->real_cpu_cnt++;
+		if (!prefix)
+			continue;
+		hdr->lc_vec[hdr->cpu_cnt] = prefix;
+		hdr->cpu_cnt++;
+	}
+}
+
+/*
+ * Provide IPL parameter information block from either HSA or memory
+ * for future reipl
+ */
+static int __init zcore_reipl_init(void)
+{
+	struct ipib_info ipib_info;
+	int rc;
+
+	rc = memcpy_hsa_kernel(&ipib_info, __LC_DUMP_REIPL, sizeof(ipib_info));
+	if (rc)
+		return rc;
+	if (ipib_info.ipib == 0)
+		return 0;
+	ipl_block = (void *) __get_free_page(GFP_KERNEL);
+	if (!ipl_block)
+		return -ENOMEM;
+	if (ipib_info.ipib < sclp_get_hsa_size())
+		rc = memcpy_hsa_kernel(ipl_block, ipib_info.ipib, PAGE_SIZE);
+	else
+		rc = memcpy_real(ipl_block, (void *) ipib_info.ipib, PAGE_SIZE);
+	if (rc || csum_partial(ipl_block, ipl_block->hdr.len, 0) !=
+	    ipib_info.checksum) {
+		TRACE("Checksum does not match\n");
+		free_page((unsigned long) ipl_block);
+		ipl_block = NULL;
+	}
+	return 0;
+}
+
+static int __init zcore_init(void)
+{
+	unsigned long mem_size, mem_end;
+	unsigned char arch;
+	int rc;
+
+	mem_size = mem_end = 0;
+	if (ipl_info.type != IPL_TYPE_FCP_DUMP)
+		return -ENODATA;
+	if (OLDMEM_BASE)
+		return -ENODATA;
+
+	zcore_dbf = debug_register("zcore", 4, 1, 4 * sizeof(long));
+	debug_register_view(zcore_dbf, &debug_sprintf_view);
+	debug_set_level(zcore_dbf, 6);
+
+	TRACE("devno:  %x\n", ipl_info.data.fcp.dev_id.devno);
+	TRACE("wwpn:   %llx\n", (unsigned long long) ipl_info.data.fcp.wwpn);
+	TRACE("lun:    %llx\n", (unsigned long long) ipl_info.data.fcp.lun);
+
+	rc = sclp_sdias_init();
+	if (rc)
+		goto fail;
+
+	rc = check_sdias();
+	if (rc)
+		goto fail;
+	hsa_available = 1;
+
+	rc = memcpy_hsa_kernel(&arch, __LC_AR_MODE_ID, 1);
+	if (rc)
+		goto fail;
+
+#ifdef CONFIG_64BIT
+	if (arch == ARCH_S390) {
+		pr_alert("The 64-bit dump tool cannot be used for a "
+			 "32-bit system\n");
+		rc = -EINVAL;
+		goto fail;
+	}
+#else /* CONFIG_64BIT */
+	if (arch == ARCH_S390X) {
+		pr_alert("The 32-bit dump tool cannot be used for a "
+			 "64-bit system\n");
+		rc = -EINVAL;
+		goto fail;
+	}
+#endif /* CONFIG_64BIT */
+
+	rc = get_mem_info(&mem_size, &mem_end);
+	if (rc)
+		goto fail;
+
+	rc = sys_info_init(arch, mem_end);
+	if (rc)
+		goto fail;
+	zcore_header_init(arch, &zcore_header, mem_size);
+
+	rc = zcore_reipl_init();
+	if (rc)
+		goto fail;
+
+	zcore_dir = debugfs_create_dir("zcore" , NULL);
+	if (!zcore_dir) {
+		rc = -ENOMEM;
+		goto fail;
+	}
+	zcore_file = debugfs_create_file("mem", S_IRUSR, zcore_dir, NULL,
+					 &zcore_fops);
+	if (!zcore_file) {
+		rc = -ENOMEM;
+		goto fail_dir;
+	}
+	zcore_memmap_file = debugfs_create_file("memmap", S_IRUSR, zcore_dir,
+						NULL, &zcore_memmap_fops);
+	if (!zcore_memmap_file) {
+		rc = -ENOMEM;
+		goto fail_file;
+	}
+	zcore_reipl_file = debugfs_create_file("reipl", S_IRUSR, zcore_dir,
+						NULL, &zcore_reipl_fops);
+	if (!zcore_reipl_file) {
+		rc = -ENOMEM;
+		goto fail_memmap_file;
+	}
+	zcore_hsa_file = debugfs_create_file("hsa", S_IRUSR|S_IWUSR, zcore_dir,
+					     NULL, &zcore_hsa_fops);
+	if (!zcore_hsa_file) {
+		rc = -ENOMEM;
+		goto fail_reipl_file;
+	}
+	return 0;
+
+fail_reipl_file:
+	debugfs_remove(zcore_reipl_file);
+fail_memmap_file:
+	debugfs_remove(zcore_memmap_file);
+fail_file:
+	debugfs_remove(zcore_file);
+fail_dir:
+	debugfs_remove(zcore_dir);
+fail:
+	diag308(DIAG308_REL_HSA, NULL);
+	return rc;
+}
+
+static void __exit zcore_exit(void)
+{
+	debug_unregister(zcore_dbf);
+	sclp_sdias_exit();
+	free_page((unsigned long) ipl_block);
+	debugfs_remove(zcore_hsa_file);
+	debugfs_remove(zcore_reipl_file);
+	debugfs_remove(zcore_memmap_file);
+	debugfs_remove(zcore_file);
+	debugfs_remove(zcore_dir);
+	diag308(DIAG308_REL_HSA, NULL);
+}
+
+MODULE_AUTHOR("Copyright IBM Corp. 2003,2008");
+MODULE_DESCRIPTION("zcore module for zfcpdump support");
+MODULE_LICENSE("GPL");
+
+subsys_initcall(zcore_init);
+module_exit(zcore_exit);

drivers/s390/cio/Makefile

@@ -0,0 +1,16 @@
+#
+# Makefile for the S/390 common i/o drivers
+#
+
+obj-y += airq.o blacklist.o chsc.o cio.o css.o chp.o idset.o isc.o \
+	fcx.o itcw.o crw.o ccwreq.o
+ccw_device-objs += device.o device_fsm.o device_ops.o
+ccw_device-objs += device_id.o device_pgid.o device_status.o
+obj-y += ccw_device.o cmf.o
+obj-$(CONFIG_CHSC_SCH) += chsc_sch.o
+obj-$(CONFIG_EADM_SCH) += eadm_sch.o
+obj-$(CONFIG_SCM_BUS) += scm.o
+obj-$(CONFIG_CCWGROUP) += ccwgroup.o
+
+qdio-objs := qdio_main.o qdio_thinint.o qdio_debug.o qdio_setup.o
+obj-$(CONFIG_QDIO) += qdio.o

drivers/s390/cio/airq.c

@@ -0,0 +1,275 @@
+/*
+ *    Support for adapter interruptions
+ *
+ *    Copyright IBM Corp. 1999, 2007
+ *    Author(s): Ingo Adlung <adlung@de.ibm.com>
+ *		 Cornelia Huck <cornelia.huck@de.ibm.com>
+ *		 Arnd Bergmann <arndb@de.ibm.com>
+ *		 Peter Oberparleiter <peter.oberparleiter@de.ibm.com>
+ */
+
+#include <linux/init.h>
+#include <linux/irq.h>
+#include <linux/kernel_stat.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/rculist.h>
+#include <linux/slab.h>
+
+#include <asm/airq.h>
+#include <asm/isc.h>
+
+#include "cio.h"
+#include "cio_debug.h"
+#include "ioasm.h"
+
+static DEFINE_SPINLOCK(airq_lists_lock);
+static struct hlist_head airq_lists[MAX_ISC+1];
+
+/**
+ * register_adapter_interrupt() - register adapter interrupt handler
+ * @airq: pointer to adapter interrupt descriptor
+ *
+ * Returns 0 on success, or -EINVAL.
+ */
+int register_adapter_interrupt(struct airq_struct *airq)
+{
+	char dbf_txt[32];
+
+	if (!airq->handler || airq->isc > MAX_ISC)
+		return -EINVAL;
+	if (!airq->lsi_ptr) {
+		airq->lsi_ptr = kzalloc(1, GFP_KERNEL);
+		if (!airq->lsi_ptr)
+			return -ENOMEM;
+		airq->flags |= AIRQ_PTR_ALLOCATED;
+	}
+	if (!airq->lsi_mask)
+		airq->lsi_mask = 0xff;
+	snprintf(dbf_txt, sizeof(dbf_txt), "rairq:%p", airq);
+	CIO_TRACE_EVENT(4, dbf_txt);
+	isc_register(airq->isc);
+	spin_lock(&airq_lists_lock);
+	hlist_add_head_rcu(&airq->list, &airq_lists[airq->isc]);
+	spin_unlock(&airq_lists_lock);
+	return 0;
+}
+EXPORT_SYMBOL(register_adapter_interrupt);
+
+/**
+ * unregister_adapter_interrupt - unregister adapter interrupt handler
+ * @airq: pointer to adapter interrupt descriptor
+ */
+void unregister_adapter_interrupt(struct airq_struct *airq)
+{
+	char dbf_txt[32];
+
+	if (hlist_unhashed(&airq->list))
+		return;
+	snprintf(dbf_txt, sizeof(dbf_txt), "urairq:%p", airq);
+	CIO_TRACE_EVENT(4, dbf_txt);
+	spin_lock(&airq_lists_lock);
+	hlist_del_rcu(&airq->list);
+	spin_unlock(&airq_lists_lock);
+	synchronize_rcu();
+	isc_unregister(airq->isc);
+	if (airq->flags & AIRQ_PTR_ALLOCATED) {
+		kfree(airq->lsi_ptr);
+		airq->lsi_ptr = NULL;
+		airq->flags &= ~AIRQ_PTR_ALLOCATED;
+	}
+}
+EXPORT_SYMBOL(unregister_adapter_interrupt);
+
+static irqreturn_t do_airq_interrupt(int irq, void *dummy)
+{
+	struct tpi_info *tpi_info;
+	struct airq_struct *airq;
+	struct hlist_head *head;
+
+	set_cpu_flag(CIF_NOHZ_DELAY);
+	tpi_info = (struct tpi_info *) &get_irq_regs()->int_code;
+	head = &airq_lists[tpi_info->isc];
+	rcu_read_lock();
+	hlist_for_each_entry_rcu(airq, head, list)
+		if ((*airq->lsi_ptr & airq->lsi_mask) != 0)
+			airq->handler(airq);
+	rcu_read_unlock();
+
+	return IRQ_HANDLED;
+}
+
+static struct irqaction airq_interrupt = {
+	.name	 = "AIO",
+	.handler = do_airq_interrupt,
+};
+
+void __init init_airq_interrupts(void)
+{
+	irq_set_chip_and_handler(THIN_INTERRUPT,
+				 &dummy_irq_chip, handle_percpu_irq);
+	setup_irq(THIN_INTERRUPT, &airq_interrupt);
+}
+
+/**
+ * airq_iv_create - create an interrupt vector
+ * @bits: number of bits in the interrupt vector
+ * @flags: allocation flags
+ *
+ * Returns a pointer to an interrupt vector structure
+ */
+struct airq_iv *airq_iv_create(unsigned long bits, unsigned long flags)
+{
+	struct airq_iv *iv;
+	unsigned long size;
+
+	iv = kzalloc(sizeof(*iv), GFP_KERNEL);
+	if (!iv)
+		goto out;
+	iv->bits = bits;
+	size = BITS_TO_LONGS(bits) * sizeof(unsigned long);
+	iv->vector = kzalloc(size, GFP_KERNEL);
+	if (!iv->vector)
+		goto out_free;
+	if (flags & AIRQ_IV_ALLOC) {
+		iv->avail = kmalloc(size, GFP_KERNEL);
+		if (!iv->avail)
+			goto out_free;
+		memset(iv->avail, 0xff, size);
+		iv->end = 0;
+	} else
+		iv->end = bits;
+	if (flags & AIRQ_IV_BITLOCK) {
+		iv->bitlock = kzalloc(size, GFP_KERNEL);
+		if (!iv->bitlock)
+			goto out_free;
+	}
+	if (flags & AIRQ_IV_PTR) {
+		size = bits * sizeof(unsigned long);
+		iv->ptr = kzalloc(size, GFP_KERNEL);
+		if (!iv->ptr)
+			goto out_free;
+	}
+	if (flags & AIRQ_IV_DATA) {
+		size = bits * sizeof(unsigned int);
+		iv->data = kzalloc(size, GFP_KERNEL);
+		if (!iv->data)
+			goto out_free;
+	}
+	spin_lock_init(&iv->lock);
+	return iv;
+
+out_free:
+	kfree(iv->ptr);
+	kfree(iv->bitlock);
+	kfree(iv->avail);
+	kfree(iv->vector);
+	kfree(iv);
+out:
+	return NULL;
+}
+EXPORT_SYMBOL(airq_iv_create);
+
+/**
+ * airq_iv_release - release an interrupt vector
+ * @iv: pointer to interrupt vector structure
+ */
+void airq_iv_release(struct airq_iv *iv)
+{
+	kfree(iv->data);
+	kfree(iv->ptr);
+	kfree(iv->bitlock);
+	kfree(iv->vector);
+	kfree(iv->avail);
+	kfree(iv);
+}
+EXPORT_SYMBOL(airq_iv_release);
+
+/**
+ * airq_iv_alloc - allocate irq bits from an interrupt vector
+ * @iv: pointer to an interrupt vector structure
+ * @num: number of consecutive irq bits to allocate
+ *
+ * Returns the bit number of the first irq in the allocated block of irqs,
+ * or -1UL if no bit is available or the AIRQ_IV_ALLOC flag has not been
+ * specified
+ */
+unsigned long airq_iv_alloc(struct airq_iv *iv, unsigned long num)
+{
+	unsigned long bit, i, flags;
+
+	if (!iv->avail || num == 0)
+		return -1UL;
+	spin_lock_irqsave(&iv->lock, flags);
+	bit = find_first_bit_inv(iv->avail, iv->bits);
+	while (bit + num <= iv->bits) {
+		for (i = 1; i < num; i++)
+			if (!test_bit_inv(bit + i, iv->avail))
+				break;
+		if (i >= num) {
+			/* Found a suitable block of irqs */
+			for (i = 0; i < num; i++)
+				clear_bit_inv(bit + i, iv->avail);
+			if (bit + num >= iv->end)
+				iv->end = bit + num + 1;
+			break;
+		}
+		bit = find_next_bit_inv(iv->avail, iv->bits, bit + i + 1);
+	}
+	if (bit + num > iv->bits)
+		bit = -1UL;
+	spin_unlock_irqrestore(&iv->lock, flags);
+	return bit;
+}
+EXPORT_SYMBOL(airq_iv_alloc);
+
+/**
+ * airq_iv_free - free irq bits of an interrupt vector
+ * @iv: pointer to interrupt vector structure
+ * @bit: number of the first irq bit to free
+ * @num: number of consecutive irq bits to free
+ */
+void airq_iv_free(struct airq_iv *iv, unsigned long bit, unsigned long num)
+{
+	unsigned long i, flags;
+
+	if (!iv->avail || num == 0)
+		return;
+	spin_lock_irqsave(&iv->lock, flags);
+	for (i = 0; i < num; i++) {
+		/* Clear (possibly left over) interrupt bit */
+		clear_bit_inv(bit + i, iv->vector);
+		/* Make the bit positions available again */
+		set_bit_inv(bit + i, iv->avail);
+	}
+	if (bit + num >= iv->end) {
+		/* Find new end of bit-field */
+		while (iv->end > 0 && !test_bit_inv(iv->end - 1, iv->avail))
+			iv->end--;
+	}
+	spin_unlock_irqrestore(&iv->lock, flags);
+}
+EXPORT_SYMBOL(airq_iv_free);
+
+/**
+ * airq_iv_scan - scan interrupt vector for non-zero bits
+ * @iv: pointer to interrupt vector structure
+ * @start: bit number to start the search
+ * @end: bit number to end the search
+ *
+ * Returns the bit number of the next non-zero interrupt bit, or
+ * -1UL if the scan completed without finding any more any non-zero bits.
+ */
+unsigned long airq_iv_scan(struct airq_iv *iv, unsigned long start,
+			   unsigned long end)
+{
+	unsigned long bit;
+
+	/* Find non-zero bit starting from 'ivs->next'. */
+	bit = find_next_bit_inv(iv->vector, end, start);
+	if (bit >= end)
+		return -1UL;
+	clear_bit_inv(bit, iv->vector);
+	return bit;
+}
+EXPORT_SYMBOL(airq_iv_scan);

drivers/s390/cio/blacklist.c

@@ -0,0 +1,421 @@
+/*
+ *   S/390 common I/O routines -- blacklisting of specific devices
+ *
+ *    Copyright IBM Corp. 1999, 2013
+ *    Author(s): Ingo Adlung (adlung@de.ibm.com)
+ *		 Cornelia Huck (cornelia.huck@de.ibm.com)
+ *		 Arnd Bergmann (arndb@de.ibm.com)
+ */
+
+#define KMSG_COMPONENT "cio"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/init.h>
+#include <linux/vmalloc.h>
+#include <linux/proc_fs.h>
+#include <linux/seq_file.h>
+#include <linux/ctype.h>
+#include <linux/device.h>
+
+#include <asm/uaccess.h>
+#include <asm/cio.h>
+#include <asm/ipl.h>
+
+#include "blacklist.h"
+#include "cio.h"
+#include "cio_debug.h"
+#include "css.h"
+#include "device.h"
+
+/*
+ * "Blacklisting" of certain devices:
+ * Device numbers given in the commandline as cio_ignore=... won't be known
+ * to Linux.
+ *
+ * These can be single devices or ranges of devices
+ */
+
+/* 65536 bits for each set to indicate if a devno is blacklisted or not */
+#define __BL_DEV_WORDS ((__MAX_SUBCHANNEL + (8*sizeof(long) - 1)) / \
+			 (8*sizeof(long)))
+static unsigned long bl_dev[__MAX_SSID + 1][__BL_DEV_WORDS];
+typedef enum {add, free} range_action;
+
+/*
+ * Function: blacklist_range
+ * (Un-)blacklist the devices from-to
+ */
+static int blacklist_range(range_action action, unsigned int from_ssid,
+			   unsigned int to_ssid, unsigned int from,
+			   unsigned int to, int msgtrigger)
+{
+	if ((from_ssid > to_ssid) || ((from_ssid == to_ssid) && (from > to))) {
+		if (msgtrigger)
+			pr_warning("0.%x.%04x to 0.%x.%04x is not a valid "
+				   "range for cio_ignore\n", from_ssid, from,
+				   to_ssid, to);
+
+		return 1;
+	}
+
+	while ((from_ssid < to_ssid) || ((from_ssid == to_ssid) &&
+	       (from <= to))) {
+		if (action == add)
+			set_bit(from, bl_dev[from_ssid]);
+		else
+			clear_bit(from, bl_dev[from_ssid]);
+		from++;
+		if (from > __MAX_SUBCHANNEL) {
+			from_ssid++;
+			from = 0;
+		}
+	}
+
+	return 0;
+}
+
+static int pure_hex(char **cp, unsigned int *val, int min_digit,
+		    int max_digit, int max_val)
+{
+	int diff;
+
+	diff = 0;
+	*val = 0;
+
+	while (diff <= max_digit) {
+		int value = hex_to_bin(**cp);
+
+		if (value < 0)
+			break;
+		*val = *val * 16 + value;
+		(*cp)++;
+		diff++;
+	}
+
+	if ((diff < min_digit) || (diff > max_digit) || (*val > max_val))
+		return 1;
+
+	return 0;
+}
+
+static int parse_busid(char *str, unsigned int *cssid, unsigned int *ssid,
+		       unsigned int *devno, int msgtrigger)
+{
+	char *str_work;
+	int val, rc, ret;
+
+	rc = 1;
+
+	if (*str == '\0')
+		goto out;
+
+	/* old style */
+	str_work = str;
+	val = simple_strtoul(str, &str_work, 16);
+
+	if (*str_work == '\0') {
+		if (val <= __MAX_SUBCHANNEL) {
+			*devno = val;
+			*ssid = 0;
+			*cssid = 0;
+			rc = 0;
+		}
+		goto out;
+	}
+
+	/* new style */
+	str_work = str;
+	ret = pure_hex(&str_work, cssid, 1, 2, __MAX_CSSID);
+	if (ret || (str_work[0] != '.'))
+		goto out;
+	str_work++;
+	ret = pure_hex(&str_work, ssid, 1, 1, __MAX_SSID);
+	if (ret || (str_work[0] != '.'))
+		goto out;
+	str_work++;
+	ret = pure_hex(&str_work, devno, 4, 4, __MAX_SUBCHANNEL);
+	if (ret || (str_work[0] != '\0'))
+		goto out;
+
+	rc = 0;
+out:
+	if (rc && msgtrigger)
+		pr_warning("%s is not a valid device for the cio_ignore "
+			   "kernel parameter\n", str);
+
+	return rc;
+}
+
+static int blacklist_parse_parameters(char *str, range_action action,
+				      int msgtrigger)
+{
+	unsigned int from_cssid, to_cssid, from_ssid, to_ssid, from, to;
+	int rc, totalrc;
+	char *parm;
+	range_action ra;
+
+	totalrc = 0;
+
+	while ((parm = strsep(&str, ","))) {
+		rc = 0;
+		ra = action;
+		if (*parm == '!') {
+			if (ra == add)
+				ra = free;
+			else
+				ra = add;
+			parm++;
+		}
+		if (strcmp(parm, "all") == 0) {
+			from_cssid = 0;
+			from_ssid = 0;
+			from = 0;
+			to_cssid = __MAX_CSSID;
+			to_ssid = __MAX_SSID;
+			to = __MAX_SUBCHANNEL;
+		} else if (strcmp(parm, "ipldev") == 0) {
+			if (ipl_info.type == IPL_TYPE_CCW) {
+				from_cssid = 0;
+				from_ssid = ipl_info.data.ccw.dev_id.ssid;
+				from = ipl_info.data.ccw.dev_id.devno;
+			} else if (ipl_info.type == IPL_TYPE_FCP ||
+				   ipl_info.type == IPL_TYPE_FCP_DUMP) {
+				from_cssid = 0;
+				from_ssid = ipl_info.data.fcp.dev_id.ssid;
+				from = ipl_info.data.fcp.dev_id.devno;
+			} else {
+				continue;
+			}
+			to_cssid = from_cssid;
+			to_ssid = from_ssid;
+			to = from;
+		} else if (strcmp(parm, "condev") == 0) {
+			if (console_devno == -1)
+				continue;
+
+			from_cssid = to_cssid = 0;
+			from_ssid = to_ssid = 0;
+			from = to = console_devno;
+		} else {
+			rc = parse_busid(strsep(&parm, "-"), &from_cssid,
+					 &from_ssid, &from, msgtrigger);
+			if (!rc) {
+				if (parm != NULL)
+					rc = parse_busid(parm, &to_cssid,
+							 &to_ssid, &to,
+							 msgtrigger);
+				else {
+					to_cssid = from_cssid;
+					to_ssid = from_ssid;
+					to = from;
+				}
+			}
+		}
+		if (!rc) {
+			rc = blacklist_range(ra, from_ssid, to_ssid, from, to,
+					     msgtrigger);
+			if (rc)
+				totalrc = -EINVAL;
+		} else
+			totalrc = -EINVAL;
+	}
+
+	return totalrc;
+}
+
+static int __init
+blacklist_setup (char *str)
+{
+	CIO_MSG_EVENT(6, "Reading blacklist parameters\n");
+	if (blacklist_parse_parameters(str, add, 1))
+		return 0;
+	return 1;
+}
+
+__setup ("cio_ignore=", blacklist_setup);
+
+/* Checking if devices are blacklisted */
+
+/*
+ * Function: is_blacklisted
+ * Returns 1 if the given devicenumber can be found in the blacklist,
+ * otherwise 0.
+ * Used by validate_subchannel()
+ */
+int
+is_blacklisted (int ssid, int devno)
+{
+	return test_bit (devno, bl_dev[ssid]);
+}
+
+#ifdef CONFIG_PROC_FS
+/*
+ * Function: blacklist_parse_proc_parameters
+ * parse the stuff which is piped to /proc/cio_ignore
+ */
+static int blacklist_parse_proc_parameters(char *buf)
+{
+	int rc;
+	char *parm;
+
+	parm = strsep(&buf, " ");
+
+	if (strcmp("free", parm) == 0) {
+		rc = blacklist_parse_parameters(buf, free, 0);
+		css_schedule_eval_all_unreg(0);
+	} else if (strcmp("add", parm) == 0)
+		rc = blacklist_parse_parameters(buf, add, 0);
+	else if (strcmp("purge", parm) == 0)
+		return ccw_purge_blacklisted();
+	else
+		return -EINVAL;
+
+
+	return rc;
+}
+
+/* Iterator struct for all devices. */
+struct ccwdev_iter {
+	int devno;
+	int ssid;
+	int in_range;
+};
+
+static void *
+cio_ignore_proc_seq_start(struct seq_file *s, loff_t *offset)
+{
+	struct ccwdev_iter *iter = s->private;
+
+	if (*offset >= (__MAX_SUBCHANNEL + 1) * (__MAX_SSID + 1))
+		return NULL;
+	memset(iter, 0, sizeof(*iter));
+	iter->ssid = *offset / (__MAX_SUBCHANNEL + 1);
+	iter->devno = *offset % (__MAX_SUBCHANNEL + 1);
+	return iter;
+}
+
+static void
+cio_ignore_proc_seq_stop(struct seq_file *s, void *it)
+{
+}
+
+static void *
+cio_ignore_proc_seq_next(struct seq_file *s, void *it, loff_t *offset)
+{
+	struct ccwdev_iter *iter;
+
+	if (*offset >= (__MAX_SUBCHANNEL + 1) * (__MAX_SSID + 1))
+		return NULL;
+	iter = it;
+	if (iter->devno == __MAX_SUBCHANNEL) {
+		iter->devno = 0;
+		iter->ssid++;
+		if (iter->ssid > __MAX_SSID)
+			return NULL;
+	} else
+		iter->devno++;
+	(*offset)++;
+	return iter;
+}
+
+static int
+cio_ignore_proc_seq_show(struct seq_file *s, void *it)
+{
+	struct ccwdev_iter *iter;
+
+	iter = it;
+	if (!is_blacklisted(iter->ssid, iter->devno))
+		/* Not blacklisted, nothing to output. */
+		return 0;
+	if (!iter->in_range) {
+		/* First device in range. */
+		if ((iter->devno == __MAX_SUBCHANNEL) ||
+		    !is_blacklisted(iter->ssid, iter->devno + 1))
+			/* Singular device. */
+			return seq_printf(s, "0.%x.%04x\n",
+					  iter->ssid, iter->devno);
+		iter->in_range = 1;
+		return seq_printf(s, "0.%x.%04x-", iter->ssid, iter->devno);
+	}
+	if ((iter->devno == __MAX_SUBCHANNEL) ||
+	    !is_blacklisted(iter->ssid, iter->devno + 1)) {
+		/* Last device in range. */
+		iter->in_range = 0;
+		return seq_printf(s, "0.%x.%04x\n", iter->ssid, iter->devno);
+	}
+	return 0;
+}
+
+static ssize_t
+cio_ignore_write(struct file *file, const char __user *user_buf,
+		 size_t user_len, loff_t *offset)
+{
+	char *buf;
+	ssize_t rc, ret, i;
+
+	if (*offset)
+		return -EINVAL;
+	if (user_len > 65536)
+		user_len = 65536;
+	buf = vzalloc(user_len + 1); /* maybe better use the stack? */
+	if (buf == NULL)
+		return -ENOMEM;
+
+	if (strncpy_from_user (buf, user_buf, user_len) < 0) {
+		rc = -EFAULT;
+		goto out_free;
+	}
+
+	i = user_len - 1;
+	while ((i >= 0) && (isspace(buf[i]) || (buf[i] == 0))) {
+		buf[i] = '\0';
+		i--;
+	}
+	ret = blacklist_parse_proc_parameters(buf);
+	if (ret)
+		rc = ret;
+	else
+		rc = user_len;
+
+out_free:
+	vfree (buf);
+	return rc;
+}
+
+static const struct seq_operations cio_ignore_proc_seq_ops = {
+	.start = cio_ignore_proc_seq_start,
+	.stop  = cio_ignore_proc_seq_stop,
+	.next  = cio_ignore_proc_seq_next,
+	.show  = cio_ignore_proc_seq_show,
+};
+
+static int
+cio_ignore_proc_open(struct inode *inode, struct file *file)
+{
+	return seq_open_private(file, &cio_ignore_proc_seq_ops,
+				sizeof(struct ccwdev_iter));
+}
+
+static const struct file_operations cio_ignore_proc_fops = {
+	.open    = cio_ignore_proc_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release_private,
+	.write   = cio_ignore_write,
+};
+
+static int
+cio_ignore_proc_init (void)
+{
+	struct proc_dir_entry *entry;
+
+	entry = proc_create("cio_ignore", S_IFREG | S_IRUGO | S_IWUSR, NULL,
+			    &cio_ignore_proc_fops);
+	if (!entry)
+		return -ENOENT;
+	return 0;
+}
+
+__initcall (cio_ignore_proc_init);
+
+#endif /* CONFIG_PROC_FS */

drivers/s390/cio/blacklist.h

@@ -0,0 +1,6 @@
+#ifndef S390_BLACKLIST_H
+#define S390_BLACKLIST_H
+
+extern int is_blacklisted (int ssid, int devno);
+
+#endif

drivers/s390/cio/ccwgroup.c

@@ -0,0 +1,641 @@
+/*
+ *  bus driver for ccwgroup
+ *
+ *  Copyright IBM Corp. 2002, 2012
+ *
+ *  Author(s): Arnd Bergmann (arndb@de.ibm.com)
+ *	       Cornelia Huck (cornelia.huck@de.ibm.com)
+ */
+#include <linux/module.h>
+#include <linux/errno.h>
+#include <linux/slab.h>
+#include <linux/list.h>
+#include <linux/device.h>
+#include <linux/init.h>
+#include <linux/ctype.h>
+#include <linux/dcache.h>
+
+#include <asm/cio.h>
+#include <asm/ccwdev.h>
+#include <asm/ccwgroup.h>
+
+#include "device.h"
+
+#define CCW_BUS_ID_SIZE		10
+
+/* In Linux 2.4, we had a channel device layer called "chandev"
+ * that did all sorts of obscure stuff for networking devices.
+ * This is another driver that serves as a replacement for just
+ * one of its functions, namely the translation of single subchannels
+ * to devices that use multiple subchannels.
+ */
+
+static struct bus_type ccwgroup_bus_type;
+
+static void __ccwgroup_remove_symlinks(struct ccwgroup_device *gdev)
+{
+	int i;
+	char str[8];
+
+	for (i = 0; i < gdev->count; i++) {
+		sprintf(str, "cdev%d", i);
+		sysfs_remove_link(&gdev->dev.kobj, str);
+		sysfs_remove_link(&gdev->cdev[i]->dev.kobj, "group_device");
+	}
+}
+
+/*
+ * Remove references from ccw devices to ccw group device and from
+ * ccw group device to ccw devices.
+ */
+static void __ccwgroup_remove_cdev_refs(struct ccwgroup_device *gdev)
+{
+	struct ccw_device *cdev;
+	int i;
+
+	for (i = 0; i < gdev->count; i++) {
+		cdev = gdev->cdev[i];
+		if (!cdev)
+			continue;
+		spin_lock_irq(cdev->ccwlock);
+		dev_set_drvdata(&cdev->dev, NULL);
+		spin_unlock_irq(cdev->ccwlock);
+		gdev->cdev[i] = NULL;
+		put_device(&cdev->dev);
+	}
+}
+
+/**
+ * ccwgroup_set_online() - enable a ccwgroup device
+ * @gdev: target ccwgroup device
+ *
+ * This function attempts to put the ccwgroup device into the online state.
+ * Returns:
+ *  %0 on success and a negative error value on failure.
+ */
+int ccwgroup_set_online(struct ccwgroup_device *gdev)
+{
+	struct ccwgroup_driver *gdrv = to_ccwgroupdrv(gdev->dev.driver);
+	int ret = -EINVAL;
+
+	if (atomic_cmpxchg(&gdev->onoff, 0, 1) != 0)
+		return -EAGAIN;
+	if (gdev->state == CCWGROUP_ONLINE)
+		goto out;
+	if (gdrv->set_online)
+		ret = gdrv->set_online(gdev);
+	if (ret)
+		goto out;
+
+	gdev->state = CCWGROUP_ONLINE;
+out:
+	atomic_set(&gdev->onoff, 0);
+	return ret;
+}
+EXPORT_SYMBOL(ccwgroup_set_online);
+
+/**
+ * ccwgroup_set_offline() - disable a ccwgroup device
+ * @gdev: target ccwgroup device
+ *
+ * This function attempts to put the ccwgroup device into the offline state.
+ * Returns:
+ *  %0 on success and a negative error value on failure.
+ */
+int ccwgroup_set_offline(struct ccwgroup_device *gdev)
+{
+	struct ccwgroup_driver *gdrv = to_ccwgroupdrv(gdev->dev.driver);
+	int ret = -EINVAL;
+
+	if (atomic_cmpxchg(&gdev->onoff, 0, 1) != 0)
+		return -EAGAIN;
+	if (gdev->state == CCWGROUP_OFFLINE)
+		goto out;
+	if (gdrv->set_offline)
+		ret = gdrv->set_offline(gdev);
+	if (ret)
+		goto out;
+
+	gdev->state = CCWGROUP_OFFLINE;
+out:
+	atomic_set(&gdev->onoff, 0);
+	return ret;
+}
+EXPORT_SYMBOL(ccwgroup_set_offline);
+
+static ssize_t ccwgroup_online_store(struct device *dev,
+				     struct device_attribute *attr,
+				     const char *buf, size_t count)
+{
+	struct ccwgroup_device *gdev = to_ccwgroupdev(dev);
+	unsigned long value;
+	int ret;
+
+	device_lock(dev);
+	if (!dev->driver) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	ret = kstrtoul(buf, 0, &value);
+	if (ret)
+		goto out;
+
+	if (value == 1)
+		ret = ccwgroup_set_online(gdev);
+	else if (value == 0)
+		ret = ccwgroup_set_offline(gdev);
+	else
+		ret = -EINVAL;
+out:
+	device_unlock(dev);
+	return (ret == 0) ? count : ret;
+}
+
+static ssize_t ccwgroup_online_show(struct device *dev,
+				    struct device_attribute *attr,
+				    char *buf)
+{
+	struct ccwgroup_device *gdev = to_ccwgroupdev(dev);
+	int online;
+
+	online = (gdev->state == CCWGROUP_ONLINE) ? 1 : 0;
+
+	return scnprintf(buf, PAGE_SIZE, "%d\n", online);
+}
+
+/*
+ * Provide an 'ungroup' attribute so the user can remove group devices no
+ * longer needed or accidentially created. Saves memory :)
+ */
+static void ccwgroup_ungroup(struct ccwgroup_device *gdev)
+{
+	mutex_lock(&gdev->reg_mutex);
+	if (device_is_registered(&gdev->dev)) {
+		__ccwgroup_remove_symlinks(gdev);
+		device_unregister(&gdev->dev);
+		__ccwgroup_remove_cdev_refs(gdev);
+	}
+	mutex_unlock(&gdev->reg_mutex);
+}
+
+static ssize_t ccwgroup_ungroup_store(struct device *dev,
+				      struct device_attribute *attr,
+				      const char *buf, size_t count)
+{
+	struct ccwgroup_device *gdev = to_ccwgroupdev(dev);
+	int rc = 0;
+
+	/* Prevent concurrent online/offline processing and ungrouping. */
+	if (atomic_cmpxchg(&gdev->onoff, 0, 1) != 0)
+		return -EAGAIN;
+	if (gdev->state != CCWGROUP_OFFLINE) {
+		rc = -EINVAL;
+		goto out;
+	}
+
+	if (device_remove_file_self(dev, attr))
+		ccwgroup_ungroup(gdev);
+	else
+		rc = -ENODEV;
+out:
+	if (rc) {
+		/* Release onoff "lock" when ungrouping failed. */
+		atomic_set(&gdev->onoff, 0);
+		return rc;
+	}
+	return count;
+}
+static DEVICE_ATTR(ungroup, 0200, NULL, ccwgroup_ungroup_store);
+static DEVICE_ATTR(online, 0644, ccwgroup_online_show, ccwgroup_online_store);
+
+static struct attribute *ccwgroup_attrs[] = {
+	&dev_attr_online.attr,
+	&dev_attr_ungroup.attr,
+	NULL,
+};
+static struct attribute_group ccwgroup_attr_group = {
+	.attrs = ccwgroup_attrs,
+};
+static const struct attribute_group *ccwgroup_attr_groups[] = {
+	&ccwgroup_attr_group,
+	NULL,
+};
+
+static void ccwgroup_ungroup_workfn(struct work_struct *work)
+{
+	struct ccwgroup_device *gdev =
+		container_of(work, struct ccwgroup_device, ungroup_work);
+
+	ccwgroup_ungroup(gdev);
+	put_device(&gdev->dev);
+}
+
+static void ccwgroup_release(struct device *dev)
+{
+	kfree(to_ccwgroupdev(dev));
+}
+
+static int __ccwgroup_create_symlinks(struct ccwgroup_device *gdev)
+{
+	char str[8];
+	int i, rc;
+
+	for (i = 0; i < gdev->count; i++) {
+		rc = sysfs_create_link(&gdev->cdev[i]->dev.kobj,
+				       &gdev->dev.kobj, "group_device");
+		if (rc) {
+			for (--i; i >= 0; i--)
+				sysfs_remove_link(&gdev->cdev[i]->dev.kobj,
+						  "group_device");
+			return rc;
+		}
+	}
+	for (i = 0; i < gdev->count; i++) {
+		sprintf(str, "cdev%d", i);
+		rc = sysfs_create_link(&gdev->dev.kobj,
+				       &gdev->cdev[i]->dev.kobj, str);
+		if (rc) {
+			for (--i; i >= 0; i--) {
+				sprintf(str, "cdev%d", i);
+				sysfs_remove_link(&gdev->dev.kobj, str);
+			}
+			for (i = 0; i < gdev->count; i++)
+				sysfs_remove_link(&gdev->cdev[i]->dev.kobj,
+						  "group_device");
+			return rc;
+		}
+	}
+	return 0;
+}
+
+static int __get_next_id(const char **buf, struct ccw_dev_id *id)
+{
+	unsigned int cssid, ssid, devno;
+	int ret = 0, len;
+	char *start, *end;
+
+	start = (char *)*buf;
+	end = strchr(start, ',');
+	if (!end) {
+		/* Last entry. Strip trailing newline, if applicable. */
+		end = strchr(start, '\n');
+		if (end)
+			*end = '\0';
+		len = strlen(start) + 1;
+	} else {
+		len = end - start + 1;
+		end++;
+	}
+	if (len <= CCW_BUS_ID_SIZE) {
+		if (sscanf(start, "%2x.%1x.%04x", &cssid, &ssid, &devno) != 3)
+			ret = -EINVAL;
+	} else
+		ret = -EINVAL;
+
+	if (!ret) {
+		id->ssid = ssid;
+		id->devno = devno;
+	}
+	*buf = end;
+	return ret;
+}
+
+/**
+ * ccwgroup_create_dev() - create and register a ccw group device
+ * @parent: parent device for the new device
+ * @gdrv: driver for the new group device
+ * @num_devices: number of slave devices
+ * @buf: buffer containing comma separated bus ids of slave devices
+ *
+ * Create and register a new ccw group device as a child of @parent. Slave
+ * devices are obtained from the list of bus ids given in @buf.
+ * Returns:
+ *  %0 on success and an error code on failure.
+ * Context:
+ *  non-atomic
+ */
+int ccwgroup_create_dev(struct device *parent, struct ccwgroup_driver *gdrv,
+			int num_devices, const char *buf)
+{
+	struct ccwgroup_device *gdev;
+	struct ccw_dev_id dev_id;
+	int rc, i;
+
+	gdev = kzalloc(sizeof(*gdev) + num_devices * sizeof(gdev->cdev[0]),
+		       GFP_KERNEL);
+	if (!gdev)
+		return -ENOMEM;
+
+	atomic_set(&gdev->onoff, 0);
+	mutex_init(&gdev->reg_mutex);
+	mutex_lock(&gdev->reg_mutex);
+	INIT_WORK(&gdev->ungroup_work, ccwgroup_ungroup_workfn);
+	gdev->count = num_devices;
+	gdev->dev.bus = &ccwgroup_bus_type;
+	gdev->dev.parent = parent;
+	gdev->dev.release = ccwgroup_release;
+	device_initialize(&gdev->dev);
+
+	for (i = 0; i < num_devices && buf; i++) {
+		rc = __get_next_id(&buf, &dev_id);
+		if (rc != 0)
+			goto error;
+		gdev->cdev[i] = get_ccwdev_by_dev_id(&dev_id);
+		/*
+		 * All devices have to be of the same type in
+		 * order to be grouped.
+		 */
+		if (!gdev->cdev[i] || !gdev->cdev[i]->drv ||
+		    gdev->cdev[i]->drv != gdev->cdev[0]->drv ||
+		    gdev->cdev[i]->id.driver_info !=
+		    gdev->cdev[0]->id.driver_info) {
+			rc = -EINVAL;
+			goto error;
+		}
+		/* Don't allow a device to belong to more than one group. */
+		spin_lock_irq(gdev->cdev[i]->ccwlock);
+		if (dev_get_drvdata(&gdev->cdev[i]->dev)) {
+			spin_unlock_irq(gdev->cdev[i]->ccwlock);
+			rc = -EINVAL;
+			goto error;
+		}
+		dev_set_drvdata(&gdev->cdev[i]->dev, gdev);
+		spin_unlock_irq(gdev->cdev[i]->ccwlock);
+	}
+	/* Check for sufficient number of bus ids. */
+	if (i < num_devices) {
+		rc = -EINVAL;
+		goto error;
+	}
+	/* Check for trailing stuff. */
+	if (i == num_devices && strlen(buf) > 0) {
+		rc = -EINVAL;
+		goto error;
+	}
+
+	dev_set_name(&gdev->dev, "%s", dev_name(&gdev->cdev[0]->dev));
+	gdev->dev.groups = ccwgroup_attr_groups;
+
+	if (gdrv) {
+		gdev->dev.driver = &gdrv->driver;
+		rc = gdrv->setup ? gdrv->setup(gdev) : 0;
+		if (rc)
+			goto error;
+	}
+	rc = device_add(&gdev->dev);
+	if (rc)
+		goto error;
+	rc = __ccwgroup_create_symlinks(gdev);
+	if (rc) {
+		device_del(&gdev->dev);
+		goto error;
+	}
+	mutex_unlock(&gdev->reg_mutex);
+	return 0;
+error:
+	for (i = 0; i < num_devices; i++)
+		if (gdev->cdev[i]) {
+			spin_lock_irq(gdev->cdev[i]->ccwlock);
+			if (dev_get_drvdata(&gdev->cdev[i]->dev) == gdev)
+				dev_set_drvdata(&gdev->cdev[i]->dev, NULL);
+			spin_unlock_irq(gdev->cdev[i]->ccwlock);
+			put_device(&gdev->cdev[i]->dev);
+			gdev->cdev[i] = NULL;
+		}
+	mutex_unlock(&gdev->reg_mutex);
+	put_device(&gdev->dev);
+	return rc;
+}
+EXPORT_SYMBOL(ccwgroup_create_dev);
+
+static int ccwgroup_notifier(struct notifier_block *nb, unsigned long action,
+			     void *data)
+{
+	struct ccwgroup_device *gdev = to_ccwgroupdev(data);
+
+	if (action == BUS_NOTIFY_UNBIND_DRIVER) {
+		get_device(&gdev->dev);
+		schedule_work(&gdev->ungroup_work);
+	}
+
+	return NOTIFY_OK;
+}
+
+static struct notifier_block ccwgroup_nb = {
+	.notifier_call = ccwgroup_notifier
+};
+
+static int __init init_ccwgroup(void)
+{
+	int ret;
+
+	ret = bus_register(&ccwgroup_bus_type);
+	if (ret)
+		return ret;
+
+	ret = bus_register_notifier(&ccwgroup_bus_type, &ccwgroup_nb);
+	if (ret)
+		bus_unregister(&ccwgroup_bus_type);
+
+	return ret;
+}
+
+static void __exit cleanup_ccwgroup(void)
+{
+	bus_unregister_notifier(&ccwgroup_bus_type, &ccwgroup_nb);
+	bus_unregister(&ccwgroup_bus_type);
+}
+
+module_init(init_ccwgroup);
+module_exit(cleanup_ccwgroup);
+
+/************************** driver stuff ******************************/
+
+static int ccwgroup_remove(struct device *dev)
+{
+	struct ccwgroup_device *gdev = to_ccwgroupdev(dev);
+	struct ccwgroup_driver *gdrv = to_ccwgroupdrv(dev->driver);
+
+	if (!dev->driver)
+		return 0;
+	if (gdrv->remove)
+		gdrv->remove(gdev);
+
+	return 0;
+}
+
+static void ccwgroup_shutdown(struct device *dev)
+{
+	struct ccwgroup_device *gdev = to_ccwgroupdev(dev);
+	struct ccwgroup_driver *gdrv = to_ccwgroupdrv(dev->driver);
+
+	if (!dev->driver)
+		return;
+	if (gdrv->shutdown)
+		gdrv->shutdown(gdev);
+}
+
+static int ccwgroup_pm_prepare(struct device *dev)
+{
+	struct ccwgroup_device *gdev = to_ccwgroupdev(dev);
+	struct ccwgroup_driver *gdrv = to_ccwgroupdrv(gdev->dev.driver);
+
+	/* Fail while device is being set online/offline. */
+	if (atomic_read(&gdev->onoff))
+		return -EAGAIN;
+
+	if (!gdev->dev.driver || gdev->state != CCWGROUP_ONLINE)
+		return 0;
+
+	return gdrv->prepare ? gdrv->prepare(gdev) : 0;
+}
+
+static void ccwgroup_pm_complete(struct device *dev)
+{
+	struct ccwgroup_device *gdev = to_ccwgroupdev(dev);
+	struct ccwgroup_driver *gdrv = to_ccwgroupdrv(dev->driver);
+
+	if (!gdev->dev.driver || gdev->state != CCWGROUP_ONLINE)
+		return;
+
+	if (gdrv->complete)
+		gdrv->complete(gdev);
+}
+
+static int ccwgroup_pm_freeze(struct device *dev)
+{
+	struct ccwgroup_device *gdev = to_ccwgroupdev(dev);
+	struct ccwgroup_driver *gdrv = to_ccwgroupdrv(gdev->dev.driver);
+
+	if (!gdev->dev.driver || gdev->state != CCWGROUP_ONLINE)
+		return 0;
+
+	return gdrv->freeze ? gdrv->freeze(gdev) : 0;
+}
+
+static int ccwgroup_pm_thaw(struct device *dev)
+{
+	struct ccwgroup_device *gdev = to_ccwgroupdev(dev);
+	struct ccwgroup_driver *gdrv = to_ccwgroupdrv(gdev->dev.driver);
+
+	if (!gdev->dev.driver || gdev->state != CCWGROUP_ONLINE)
+		return 0;
+
+	return gdrv->thaw ? gdrv->thaw(gdev) : 0;
+}
+
+static int ccwgroup_pm_restore(struct device *dev)
+{
+	struct ccwgroup_device *gdev = to_ccwgroupdev(dev);
+	struct ccwgroup_driver *gdrv = to_ccwgroupdrv(gdev->dev.driver);
+
+	if (!gdev->dev.driver || gdev->state != CCWGROUP_ONLINE)
+		return 0;
+
+	return gdrv->restore ? gdrv->restore(gdev) : 0;
+}
+
+static const struct dev_pm_ops ccwgroup_pm_ops = {
+	.prepare = ccwgroup_pm_prepare,
+	.complete = ccwgroup_pm_complete,
+	.freeze = ccwgroup_pm_freeze,
+	.thaw = ccwgroup_pm_thaw,
+	.restore = ccwgroup_pm_restore,
+};
+
+static struct bus_type ccwgroup_bus_type = {
+	.name   = "ccwgroup",
+	.remove = ccwgroup_remove,
+	.shutdown = ccwgroup_shutdown,
+	.pm = &ccwgroup_pm_ops,
+};
+
+/**
+ * ccwgroup_driver_register() - register a ccw group driver
+ * @cdriver: driver to be registered
+ *
+ * This function is mainly a wrapper around driver_register().
+ */
+int ccwgroup_driver_register(struct ccwgroup_driver *cdriver)
+{
+	/* register our new driver with the core */
+	cdriver->driver.bus = &ccwgroup_bus_type;
+
+	return driver_register(&cdriver->driver);
+}
+EXPORT_SYMBOL(ccwgroup_driver_register);
+
+static int __ccwgroup_match_all(struct device *dev, void *data)
+{
+	return 1;
+}
+
+/**
+ * ccwgroup_driver_unregister() - deregister a ccw group driver
+ * @cdriver: driver to be deregistered
+ *
+ * This function is mainly a wrapper around driver_unregister().
+ */
+void ccwgroup_driver_unregister(struct ccwgroup_driver *cdriver)
+{
+	struct device *dev;
+
+	/* We don't want ccwgroup devices to live longer than their driver. */
+	while ((dev = driver_find_device(&cdriver->driver, NULL, NULL,
+					 __ccwgroup_match_all))) {
+		struct ccwgroup_device *gdev = to_ccwgroupdev(dev);
+
+		ccwgroup_ungroup(gdev);
+		put_device(dev);
+	}
+	driver_unregister(&cdriver->driver);
+}
+EXPORT_SYMBOL(ccwgroup_driver_unregister);
+
+/**
+ * ccwgroup_probe_ccwdev() - probe function for slave devices
+ * @cdev: ccw device to be probed
+ *
+ * This is a dummy probe function for ccw devices that are slave devices in
+ * a ccw group device.
+ * Returns:
+ *  always %0
+ */
+int ccwgroup_probe_ccwdev(struct ccw_device *cdev)
+{
+	return 0;
+}
+EXPORT_SYMBOL(ccwgroup_probe_ccwdev);
+
+/**
+ * ccwgroup_remove_ccwdev() - remove function for slave devices
+ * @cdev: ccw device to be removed
+ *
+ * This is a remove function for ccw devices that are slave devices in a ccw
+ * group device. It sets the ccw device offline and also deregisters the
+ * embedding ccw group device.
+ */
+void ccwgroup_remove_ccwdev(struct ccw_device *cdev)
+{
+	struct ccwgroup_device *gdev;
+
+	/* Ignore offlining errors, device is gone anyway. */
+	ccw_device_set_offline(cdev);
+	/* If one of its devices is gone, the whole group is done for. */
+	spin_lock_irq(cdev->ccwlock);
+	gdev = dev_get_drvdata(&cdev->dev);
+	if (!gdev) {
+		spin_unlock_irq(cdev->ccwlock);
+		return;
+	}
+	/* Get ccwgroup device reference for local processing. */
+	get_device(&gdev->dev);
+	spin_unlock_irq(cdev->ccwlock);
+	/* Unregister group device. */
+	ccwgroup_ungroup(gdev);
+	/* Release ccwgroup device reference for local processing. */
+	put_device(&gdev->dev);
+}
+EXPORT_SYMBOL(ccwgroup_remove_ccwdev);
+MODULE_LICENSE("GPL");

drivers/s390/cio/ccwreq.c

@@ -0,0 +1,367 @@
+/*
+ *  Handling of internal CCW device requests.
+ *
+ *    Copyright IBM Corp. 2009, 2011
+ *    Author(s): Peter Oberparleiter <peter.oberparleiter@de.ibm.com>
+ */
+
+#define KMSG_COMPONENT "cio"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/types.h>
+#include <linux/err.h>
+#include <asm/ccwdev.h>
+#include <asm/cio.h>
+
+#include "io_sch.h"
+#include "cio.h"
+#include "device.h"
+#include "cio_debug.h"
+
+/**
+ * lpm_adjust - adjust path mask
+ * @lpm: path mask to adjust
+ * @mask: mask of available paths
+ *
+ * Shift @lpm right until @lpm and @mask have at least one bit in common or
+ * until @lpm is zero. Return the resulting lpm.
+ */
+int lpm_adjust(int lpm, int mask)
+{
+	while (lpm && ((lpm & mask) == 0))
+		lpm >>= 1;
+	return lpm;
+}
+
+/*
+ * Adjust path mask to use next path and reset retry count. Return resulting
+ * path mask.
+ */
+static u16 ccwreq_next_path(struct ccw_device *cdev)
+{
+	struct ccw_request *req = &cdev->private->req;
+
+	if (!req->singlepath) {
+		req->mask = 0;
+		goto out;
+	}
+	req->retries	= req->maxretries;
+	req->mask	= lpm_adjust(req->mask >> 1, req->lpm);
+out:
+	return req->mask;
+}
+
+/*
+ * Clean up device state and report to callback.
+ */
+static void ccwreq_stop(struct ccw_device *cdev, int rc)
+{
+	struct ccw_request *req = &cdev->private->req;
+
+	if (req->done)
+		return;
+	req->done = 1;
+	ccw_device_set_timeout(cdev, 0);
+	memset(&cdev->private->irb, 0, sizeof(struct irb));
+	if (rc && rc != -ENODEV && req->drc)
+		rc = req->drc;
+	req->callback(cdev, req->data, rc);
+}
+
+/*
+ * (Re-)Start the operation until retries and paths are exhausted.
+ */
+static void ccwreq_do(struct ccw_device *cdev)
+{
+	struct ccw_request *req = &cdev->private->req;
+	struct subchannel *sch = to_subchannel(cdev->dev.parent);
+	struct ccw1 *cp = req->cp;
+	int rc = -EACCES;
+
+	while (req->mask) {
+		if (req->retries-- == 0) {
+			/* Retries exhausted, try next path. */
+			ccwreq_next_path(cdev);
+			continue;
+		}
+		/* Perform start function. */
+		memset(&cdev->private->irb, 0, sizeof(struct irb));
+		rc = cio_start(sch, cp, (u8) req->mask);
+		if (rc == 0) {
+			/* I/O started successfully. */
+			ccw_device_set_timeout(cdev, req->timeout);
+			return;
+		}
+		if (rc == -ENODEV) {
+			/* Permanent device error. */
+			break;
+		}
+		if (rc == -EACCES) {
+			/* Permant path error. */
+			ccwreq_next_path(cdev);
+			continue;
+		}
+		/* Temporary improper status. */
+		rc = cio_clear(sch);
+		if (rc)
+			break;
+		return;
+	}
+	ccwreq_stop(cdev, rc);
+}
+
+/**
+ * ccw_request_start - perform I/O request
+ * @cdev: ccw device
+ *
+ * Perform the I/O request specified by cdev->req.
+ */
+void ccw_request_start(struct ccw_device *cdev)
+{
+	struct ccw_request *req = &cdev->private->req;
+
+	if (req->singlepath) {
+		/* Try all paths twice to counter link flapping. */
+		req->mask = 0x8080;
+	} else
+		req->mask = req->lpm;
+
+	req->retries	= req->maxretries;
+	req->mask	= lpm_adjust(req->mask, req->lpm);
+	req->drc	= 0;
+	req->done	= 0;
+	req->cancel	= 0;
+	if (!req->mask)
+		goto out_nopath;
+	ccwreq_do(cdev);
+	return;
+
+out_nopath:
+	ccwreq_stop(cdev, -EACCES);
+}
+
+/**
+ * ccw_request_cancel - cancel running I/O request
+ * @cdev: ccw device
+ *
+ * Cancel the I/O request specified by cdev->req. Return non-zero if request
+ * has already finished, zero otherwise.
+ */
+int ccw_request_cancel(struct ccw_device *cdev)
+{
+	struct subchannel *sch = to_subchannel(cdev->dev.parent);
+	struct ccw_request *req = &cdev->private->req;
+	int rc;
+
+	if (req->done)
+		return 1;
+	req->cancel = 1;
+	rc = cio_clear(sch);
+	if (rc)
+		ccwreq_stop(cdev, rc);
+	return 0;
+}
+
+/*
+ * Return the status of the internal I/O started on the specified ccw device.
+ * Perform BASIC SENSE if required.
+ */
+static enum io_status ccwreq_status(struct ccw_device *cdev, struct irb *lcirb)
+{
+	struct irb *irb = &cdev->private->irb;
+	struct cmd_scsw *scsw = &irb->scsw.cmd;
+	enum uc_todo todo;
+
+	/* Perform BASIC SENSE if needed. */
+	if (ccw_device_accumulate_and_sense(cdev, lcirb))
+		return IO_RUNNING;
+	/* Check for halt/clear interrupt. */
+	if (scsw->fctl & (SCSW_FCTL_HALT_FUNC | SCSW_FCTL_CLEAR_FUNC))
+		return IO_KILLED;
+	/* Check for path error. */
+	if (scsw->cc == 3 || scsw->pno)
+		return IO_PATH_ERROR;
+	/* Handle BASIC SENSE data. */
+	if (irb->esw.esw0.erw.cons) {
+		CIO_TRACE_EVENT(2, "sensedata");
+		CIO_HEX_EVENT(2, &cdev->private->dev_id,
+			      sizeof(struct ccw_dev_id));
+		CIO_HEX_EVENT(2, &cdev->private->irb.ecw, SENSE_MAX_COUNT);
+		/* Check for command reject. */
+		if (irb->ecw[0] & SNS0_CMD_REJECT)
+			return IO_REJECTED;
+		/* Ask the driver what to do */
+		if (cdev->drv && cdev->drv->uc_handler) {
+			todo = cdev->drv->uc_handler(cdev, lcirb);
+			CIO_TRACE_EVENT(2, "uc_response");
+			CIO_HEX_EVENT(2, &todo, sizeof(todo));
+			switch (todo) {
+			case UC_TODO_RETRY:
+				return IO_STATUS_ERROR;
+			case UC_TODO_RETRY_ON_NEW_PATH:
+				return IO_PATH_ERROR;
+			case UC_TODO_STOP:
+				return IO_REJECTED;
+			default:
+				return IO_STATUS_ERROR;
+			}
+		}
+		/* Assume that unexpected SENSE data implies an error. */
+		return IO_STATUS_ERROR;
+	}
+	/* Check for channel errors. */
+	if (scsw->cstat != 0)
+		return IO_STATUS_ERROR;
+	/* Check for device errors. */
+	if (scsw->dstat & ~(DEV_STAT_CHN_END | DEV_STAT_DEV_END))
+		return IO_STATUS_ERROR;
+	/* Check for final state. */
+	if (!(scsw->dstat & DEV_STAT_DEV_END))
+		return IO_RUNNING;
+	/* Check for other improper status. */
+	if (scsw->cc == 1 && (scsw->stctl & SCSW_STCTL_ALERT_STATUS))
+		return IO_STATUS_ERROR;
+	return IO_DONE;
+}
+
+/*
+ * Log ccw request status.
+ */
+static void ccwreq_log_status(struct ccw_device *cdev, enum io_status status)
+{
+	struct ccw_request *req = &cdev->private->req;
+	struct {
+		struct ccw_dev_id dev_id;
+		u16 retries;
+		u8 lpm;
+		u8 status;
+	}  __attribute__ ((packed)) data;
+	data.dev_id	= cdev->private->dev_id;
+	data.retries	= req->retries;
+	data.lpm	= (u8) req->mask;
+	data.status	= (u8) status;
+	CIO_TRACE_EVENT(2, "reqstat");
+	CIO_HEX_EVENT(2, &data, sizeof(data));
+}
+
+/**
+ * ccw_request_handler - interrupt handler for I/O request procedure.
+ * @cdev: ccw device
+ *
+ * Handle interrupt during I/O request procedure.
+ */
+void ccw_request_handler(struct ccw_device *cdev)
+{
+	struct irb *irb = this_cpu_ptr(&cio_irb);
+	struct ccw_request *req = &cdev->private->req;
+	enum io_status status;
+	int rc = -EOPNOTSUPP;
+
+	/* Check status of I/O request. */
+	status = ccwreq_status(cdev, irb);
+	if (req->filter)
+		status = req->filter(cdev, req->data, irb, status);
+	if (status != IO_RUNNING)
+		ccw_device_set_timeout(cdev, 0);
+	if (status != IO_DONE && status != IO_RUNNING)
+		ccwreq_log_status(cdev, status);
+	switch (status) {
+	case IO_DONE:
+		break;
+	case IO_RUNNING:
+		return;
+	case IO_REJECTED:
+		goto err;
+	case IO_PATH_ERROR:
+		goto out_next_path;
+	case IO_STATUS_ERROR:
+		goto out_restart;
+	case IO_KILLED:
+		/* Check if request was cancelled on purpose. */
+		if (req->cancel) {
+			rc = -EIO;
+			goto err;
+		}
+		goto out_restart;
+	}
+	/* Check back with request initiator. */
+	if (!req->check)
+		goto out;
+	switch (req->check(cdev, req->data)) {
+	case 0:
+		break;
+	case -EAGAIN:
+		goto out_restart;
+	case -EACCES:
+		goto out_next_path;
+	default:
+		goto err;
+	}
+out:
+	ccwreq_stop(cdev, 0);
+	return;
+
+out_next_path:
+	/* Try next path and restart I/O. */
+	if (!ccwreq_next_path(cdev)) {
+		rc = -EACCES;
+		goto err;
+	}
+out_restart:
+	/* Restart. */
+	ccwreq_do(cdev);
+	return;
+err:
+	ccwreq_stop(cdev, rc);
+}
+
+
+/**
+ * ccw_request_timeout - timeout handler for I/O request procedure
+ * @cdev: ccw device
+ *
+ * Handle timeout during I/O request procedure.
+ */
+void ccw_request_timeout(struct ccw_device *cdev)
+{
+	struct subchannel *sch = to_subchannel(cdev->dev.parent);
+	struct ccw_request *req = &cdev->private->req;
+	int rc = -ENODEV, chp;
+
+	if (cio_update_schib(sch))
+		goto err;
+
+	for (chp = 0; chp < 8; chp++) {
+		if ((0x80 >> chp) & sch->schib.pmcw.lpum)
+			pr_warning("%s: No interrupt was received within %lus "
+				   "(CS=%02x, DS=%02x, CHPID=%x.%02x)\n",
+				   dev_name(&cdev->dev), req->timeout / HZ,
+				   scsw_cstat(&sch->schib.scsw),
+				   scsw_dstat(&sch->schib.scsw),
+				   sch->schid.cssid,
+				   sch->schib.pmcw.chpid[chp]);
+	}
+
+	if (!ccwreq_next_path(cdev)) {
+		/* set the final return code for this request */
+		req->drc = -ETIME;
+	}
+	rc = cio_clear(sch);
+	if (rc)
+		goto err;
+	return;
+
+err:
+	ccwreq_stop(cdev, rc);
+}
+
+/**
+ * ccw_request_notoper - notoper handler for I/O request procedure
+ * @cdev: ccw device
+ *
+ * Handle notoper during I/O request procedure.
+ */
+void ccw_request_notoper(struct ccw_device *cdev)
+{
+	ccwreq_stop(cdev, -ENODEV);
+}

drivers/s390/cio/chp.c

@@ -0,0 +1,792 @@
+/*
+ *    Copyright IBM Corp. 1999, 2010
+ *    Author(s): Cornelia Huck (cornelia.huck@de.ibm.com)
+ *		 Arnd Bergmann (arndb@de.ibm.com)
+ *		 Peter Oberparleiter <peter.oberparleiter@de.ibm.com>
+ */
+
+#include <linux/bug.h>
+#include <linux/workqueue.h>
+#include <linux/spinlock.h>
+#include <linux/export.h>
+#include <linux/sched.h>
+#include <linux/init.h>
+#include <linux/jiffies.h>
+#include <linux/wait.h>
+#include <linux/mutex.h>
+#include <linux/errno.h>
+#include <linux/slab.h>
+#include <asm/chpid.h>
+#include <asm/sclp.h>
+#include <asm/crw.h>
+
+#include "cio.h"
+#include "css.h"
+#include "ioasm.h"
+#include "cio_debug.h"
+#include "chp.h"
+
+#define to_channelpath(device) container_of(device, struct channel_path, dev)
+#define CHP_INFO_UPDATE_INTERVAL	1*HZ
+
+enum cfg_task_t {
+	cfg_none,
+	cfg_configure,
+	cfg_deconfigure
+};
+
+/* Map for pending configure tasks. */
+static enum cfg_task_t chp_cfg_task[__MAX_CSSID + 1][__MAX_CHPID + 1];
+static DEFINE_MUTEX(cfg_lock);
+static int cfg_busy;
+
+/* Map for channel-path status. */
+static struct sclp_chp_info chp_info;
+static DEFINE_MUTEX(info_lock);
+
+/* Time after which channel-path status may be outdated. */
+static unsigned long chp_info_expires;
+
+/* Workqueue to perform pending configure tasks. */
+static struct workqueue_struct *chp_wq;
+static struct work_struct cfg_work;
+
+/* Wait queue for configure completion events. */
+static wait_queue_head_t cfg_wait_queue;
+
+/* Set vary state for given chpid. */
+static void set_chp_logically_online(struct chp_id chpid, int onoff)
+{
+	chpid_to_chp(chpid)->state = onoff;
+}
+
+/* On success return 0 if channel-path is varied offline, 1 if it is varied
+ * online. Return -ENODEV if channel-path is not registered. */
+int chp_get_status(struct chp_id chpid)
+{
+	return (chpid_to_chp(chpid) ? chpid_to_chp(chpid)->state : -ENODEV);
+}
+
+/**
+ * chp_get_sch_opm - return opm for subchannel
+ * @sch: subchannel
+ *
+ * Calculate and return the operational path mask (opm) based on the chpids
+ * used by the subchannel and the status of the associated channel-paths.
+ */
+u8 chp_get_sch_opm(struct subchannel *sch)
+{
+	struct chp_id chpid;
+	int opm;
+	int i;
+
+	opm = 0;
+	chp_id_init(&chpid);
+	for (i = 0; i < 8; i++) {
+		opm <<= 1;
+		chpid.id = sch->schib.pmcw.chpid[i];
+		if (chp_get_status(chpid) != 0)
+			opm |= 1;
+	}
+	return opm;
+}
+EXPORT_SYMBOL_GPL(chp_get_sch_opm);
+
+/**
+ * chp_is_registered - check if a channel-path is registered
+ * @chpid: channel-path ID
+ *
+ * Return non-zero if a channel-path with the given chpid is registered,
+ * zero otherwise.
+ */
+int chp_is_registered(struct chp_id chpid)
+{
+	return chpid_to_chp(chpid) != NULL;
+}
+
+/*
+ * Function: s390_vary_chpid
+ * Varies the specified chpid online or offline
+ */
+static int s390_vary_chpid(struct chp_id chpid, int on)
+{
+	char dbf_text[15];
+	int status;
+
+	sprintf(dbf_text, on?"varyon%x.%02x":"varyoff%x.%02x", chpid.cssid,
+		chpid.id);
+	CIO_TRACE_EVENT(2, dbf_text);
+
+	status = chp_get_status(chpid);
+	if (!on && !status)
+		return 0;
+
+	set_chp_logically_online(chpid, on);
+	chsc_chp_vary(chpid, on);
+	return 0;
+}
+
+/*
+ * Channel measurement related functions
+ */
+static ssize_t chp_measurement_chars_read(struct file *filp,
+					  struct kobject *kobj,
+					  struct bin_attribute *bin_attr,
+					  char *buf, loff_t off, size_t count)
+{
+	struct channel_path *chp;
+	struct device *device;
+
+	device = container_of(kobj, struct device, kobj);
+	chp = to_channelpath(device);
+	if (!chp->cmg_chars)
+		return 0;
+
+	return memory_read_from_buffer(buf, count, &off,
+				chp->cmg_chars, sizeof(struct cmg_chars));
+}
+
+static struct bin_attribute chp_measurement_chars_attr = {
+	.attr = {
+		.name = "measurement_chars",
+		.mode = S_IRUSR,
+	},
+	.size = sizeof(struct cmg_chars),
+	.read = chp_measurement_chars_read,
+};
+
+static void chp_measurement_copy_block(struct cmg_entry *buf,
+				       struct channel_subsystem *css,
+				       struct chp_id chpid)
+{
+	void *area;
+	struct cmg_entry *entry, reference_buf;
+	int idx;
+
+	if (chpid.id < 128) {
+		area = css->cub_addr1;
+		idx = chpid.id;
+	} else {
+		area = css->cub_addr2;
+		idx = chpid.id - 128;
+	}
+	entry = area + (idx * sizeof(struct cmg_entry));
+	do {
+		memcpy(buf, entry, sizeof(*entry));
+		memcpy(&reference_buf, entry, sizeof(*entry));
+	} while (reference_buf.values[0] != buf->values[0]);
+}
+
+static ssize_t chp_measurement_read(struct file *filp, struct kobject *kobj,
+				    struct bin_attribute *bin_attr,
+				    char *buf, loff_t off, size_t count)
+{
+	struct channel_path *chp;
+	struct channel_subsystem *css;
+	struct device *device;
+	unsigned int size;
+
+	device = container_of(kobj, struct device, kobj);
+	chp = to_channelpath(device);
+	css = to_css(chp->dev.parent);
+
+	size = sizeof(struct cmg_entry);
+
+	/* Only allow single reads. */
+	if (off || count < size)
+		return 0;
+	chp_measurement_copy_block((struct cmg_entry *)buf, css, chp->chpid);
+	count = size;
+	return count;
+}
+
+static struct bin_attribute chp_measurement_attr = {
+	.attr = {
+		.name = "measurement",
+		.mode = S_IRUSR,
+	},
+	.size = sizeof(struct cmg_entry),
+	.read = chp_measurement_read,
+};
+
+void chp_remove_cmg_attr(struct channel_path *chp)
+{
+	device_remove_bin_file(&chp->dev, &chp_measurement_chars_attr);
+	device_remove_bin_file(&chp->dev, &chp_measurement_attr);
+}
+
+int chp_add_cmg_attr(struct channel_path *chp)
+{
+	int ret;
+
+	ret = device_create_bin_file(&chp->dev, &chp_measurement_chars_attr);
+	if (ret)
+		return ret;
+	ret = device_create_bin_file(&chp->dev, &chp_measurement_attr);
+	if (ret)
+		device_remove_bin_file(&chp->dev, &chp_measurement_chars_attr);
+	return ret;
+}
+
+/*
+ * Files for the channel path entries.
+ */
+static ssize_t chp_status_show(struct device *dev,
+			       struct device_attribute *attr, char *buf)
+{
+	struct channel_path *chp = to_channelpath(dev);
+	int status;
+
+	mutex_lock(&chp->lock);
+	status = chp->state;
+	mutex_unlock(&chp->lock);
+
+	return status ? sprintf(buf, "online\n") : sprintf(buf, "offline\n");
+}
+
+static ssize_t chp_status_write(struct device *dev,
+				struct device_attribute *attr,
+				const char *buf, size_t count)
+{
+	struct channel_path *cp = to_channelpath(dev);
+	char cmd[10];
+	int num_args;
+	int error;
+
+	num_args = sscanf(buf, "%5s", cmd);
+	if (!num_args)
+		return count;
+
+	if (!strncasecmp(cmd, "on", 2) || !strcmp(cmd, "1")) {
+		mutex_lock(&cp->lock);
+		error = s390_vary_chpid(cp->chpid, 1);
+		mutex_unlock(&cp->lock);
+	} else if (!strncasecmp(cmd, "off", 3) || !strcmp(cmd, "0")) {
+		mutex_lock(&cp->lock);
+		error = s390_vary_chpid(cp->chpid, 0);
+		mutex_unlock(&cp->lock);
+	} else
+		error = -EINVAL;
+
+	return error < 0 ? error : count;
+}
+
+static DEVICE_ATTR(status, 0644, chp_status_show, chp_status_write);
+
+static ssize_t chp_configure_show(struct device *dev,
+				  struct device_attribute *attr, char *buf)
+{
+	struct channel_path *cp;
+	int status;
+
+	cp = to_channelpath(dev);
+	status = chp_info_get_status(cp->chpid);
+	if (status < 0)
+		return status;
+
+	return snprintf(buf, PAGE_SIZE, "%d\n", status);
+}
+
+static int cfg_wait_idle(void);
+
+static ssize_t chp_configure_write(struct device *dev,
+				   struct device_attribute *attr,
+				   const char *buf, size_t count)
+{
+	struct channel_path *cp;
+	int val;
+	char delim;
+
+	if (sscanf(buf, "%d %c", &val, &delim) != 1)
+		return -EINVAL;
+	if (val != 0 && val != 1)
+		return -EINVAL;
+	cp = to_channelpath(dev);
+	chp_cfg_schedule(cp->chpid, val);
+	cfg_wait_idle();
+
+	return count;
+}
+
+static DEVICE_ATTR(configure, 0644, chp_configure_show, chp_configure_write);
+
+static ssize_t chp_type_show(struct device *dev, struct device_attribute *attr,
+			     char *buf)
+{
+	struct channel_path *chp = to_channelpath(dev);
+	u8 type;
+
+	mutex_lock(&chp->lock);
+	type = chp->desc.desc;
+	mutex_unlock(&chp->lock);
+	return sprintf(buf, "%x\n", type);
+}
+
+static DEVICE_ATTR(type, 0444, chp_type_show, NULL);
+
+static ssize_t chp_cmg_show(struct device *dev, struct device_attribute *attr,
+			    char *buf)
+{
+	struct channel_path *chp = to_channelpath(dev);
+
+	if (!chp)
+		return 0;
+	if (chp->cmg == -1) /* channel measurements not available */
+		return sprintf(buf, "unknown\n");
+	return sprintf(buf, "%x\n", chp->cmg);
+}
+
+static DEVICE_ATTR(cmg, 0444, chp_cmg_show, NULL);
+
+static ssize_t chp_shared_show(struct device *dev,
+			       struct device_attribute *attr, char *buf)
+{
+	struct channel_path *chp = to_channelpath(dev);
+
+	if (!chp)
+		return 0;
+	if (chp->shared == -1) /* channel measurements not available */
+		return sprintf(buf, "unknown\n");
+	return sprintf(buf, "%x\n", chp->shared);
+}
+
+static DEVICE_ATTR(shared, 0444, chp_shared_show, NULL);
+
+static ssize_t chp_chid_show(struct device *dev, struct device_attribute *attr,
+			     char *buf)
+{
+	struct channel_path *chp = to_channelpath(dev);
+	ssize_t rc;
+
+	mutex_lock(&chp->lock);
+	if (chp->desc_fmt1.flags & 0x10)
+		rc = sprintf(buf, "%04x\n", chp->desc_fmt1.chid);
+	else
+		rc = 0;
+	mutex_unlock(&chp->lock);
+
+	return rc;
+}
+static DEVICE_ATTR(chid, 0444, chp_chid_show, NULL);
+
+static ssize_t chp_chid_external_show(struct device *dev,
+				      struct device_attribute *attr, char *buf)
+{
+	struct channel_path *chp = to_channelpath(dev);
+	ssize_t rc;
+
+	mutex_lock(&chp->lock);
+	if (chp->desc_fmt1.flags & 0x10)
+		rc = sprintf(buf, "%x\n", chp->desc_fmt1.flags & 0x8 ? 1 : 0);
+	else
+		rc = 0;
+	mutex_unlock(&chp->lock);
+
+	return rc;
+}
+static DEVICE_ATTR(chid_external, 0444, chp_chid_external_show, NULL);
+
+static struct attribute *chp_attrs[] = {
+	&dev_attr_status.attr,
+	&dev_attr_configure.attr,
+	&dev_attr_type.attr,
+	&dev_attr_cmg.attr,
+	&dev_attr_shared.attr,
+	&dev_attr_chid.attr,
+	&dev_attr_chid_external.attr,
+	NULL,
+};
+static struct attribute_group chp_attr_group = {
+	.attrs = chp_attrs,
+};
+static const struct attribute_group *chp_attr_groups[] = {
+	&chp_attr_group,
+	NULL,
+};
+
+static void chp_release(struct device *dev)
+{
+	struct channel_path *cp;
+
+	cp = to_channelpath(dev);
+	kfree(cp);
+}
+
+/**
+ * chp_update_desc - update channel-path description
+ * @chp - channel-path
+ *
+ * Update the channel-path description of the specified channel-path.
+ * Return zero on success, non-zero otherwise.
+ */
+int chp_update_desc(struct channel_path *chp)
+{
+	int rc;
+
+	rc = chsc_determine_base_channel_path_desc(chp->chpid, &chp->desc);
+	if (rc)
+		return rc;
+
+	rc = chsc_determine_fmt1_channel_path_desc(chp->chpid, &chp->desc_fmt1);
+
+	return rc;
+}
+
+/**
+ * chp_new - register a new channel-path
+ * @chpid - channel-path ID
+ *
+ * Create and register data structure representing new channel-path. Return
+ * zero on success, non-zero otherwise.
+ */
+int chp_new(struct chp_id chpid)
+{
+	struct channel_path *chp;
+	int ret;
+
+	if (chp_is_registered(chpid))
+		return 0;
+	chp = kzalloc(sizeof(struct channel_path), GFP_KERNEL);
+	if (!chp)
+		return -ENOMEM;
+
+	/* fill in status, etc. */
+	chp->chpid = chpid;
+	chp->state = 1;
+	chp->dev.parent = &channel_subsystems[chpid.cssid]->device;
+	chp->dev.groups = chp_attr_groups;
+	chp->dev.release = chp_release;
+	mutex_init(&chp->lock);
+
+	/* Obtain channel path description and fill it in. */
+	ret = chp_update_desc(chp);
+	if (ret)
+		goto out_free;
+	if ((chp->desc.flags & 0x80) == 0) {
+		ret = -ENODEV;
+		goto out_free;
+	}
+	/* Get channel-measurement characteristics. */
+	if (css_chsc_characteristics.scmc && css_chsc_characteristics.secm) {
+		ret = chsc_get_channel_measurement_chars(chp);
+		if (ret)
+			goto out_free;
+	} else {
+		chp->cmg = -1;
+	}
+	dev_set_name(&chp->dev, "chp%x.%02x", chpid.cssid, chpid.id);
+
+	/* make it known to the system */
+	ret = device_register(&chp->dev);
+	if (ret) {
+		CIO_MSG_EVENT(0, "Could not register chp%x.%02x: %d\n",
+			      chpid.cssid, chpid.id, ret);
+		put_device(&chp->dev);
+		goto out;
+	}
+	mutex_lock(&channel_subsystems[chpid.cssid]->mutex);
+	if (channel_subsystems[chpid.cssid]->cm_enabled) {
+		ret = chp_add_cmg_attr(chp);
+		if (ret) {
+			device_unregister(&chp->dev);
+			mutex_unlock(&channel_subsystems[chpid.cssid]->mutex);
+			goto out;
+		}
+	}
+	channel_subsystems[chpid.cssid]->chps[chpid.id] = chp;
+	mutex_unlock(&channel_subsystems[chpid.cssid]->mutex);
+	goto out;
+out_free:
+	kfree(chp);
+out:
+	return ret;
+}
+
+/**
+ * chp_get_chp_desc - return newly allocated channel-path description
+ * @chpid: channel-path ID
+ *
+ * On success return a newly allocated copy of the channel-path description
+ * data associated with the given channel-path ID. Return %NULL on error.
+ */
+struct channel_path_desc *chp_get_chp_desc(struct chp_id chpid)
+{
+	struct channel_path *chp;
+	struct channel_path_desc *desc;
+
+	chp = chpid_to_chp(chpid);
+	if (!chp)
+		return NULL;
+	desc = kmalloc(sizeof(struct channel_path_desc), GFP_KERNEL);
+	if (!desc)
+		return NULL;
+
+	mutex_lock(&chp->lock);
+	memcpy(desc, &chp->desc, sizeof(struct channel_path_desc));
+	mutex_unlock(&chp->lock);
+	return desc;
+}
+
+/**
+ * chp_process_crw - process channel-path status change
+ * @crw0: channel report-word to handler
+ * @crw1: second channel-report word (always NULL)
+ * @overflow: crw overflow indication
+ *
+ * Handle channel-report-words indicating that the status of a channel-path
+ * has changed.
+ */
+static void chp_process_crw(struct crw *crw0, struct crw *crw1,
+			    int overflow)
+{
+	struct chp_id chpid;
+
+	if (overflow) {
+		css_schedule_eval_all();
+		return;
+	}
+	CIO_CRW_EVENT(2, "CRW reports slct=%d, oflw=%d, "
+		      "chn=%d, rsc=%X, anc=%d, erc=%X, rsid=%X\n",
+		      crw0->slct, crw0->oflw, crw0->chn, crw0->rsc, crw0->anc,
+		      crw0->erc, crw0->rsid);
+	/*
+	 * Check for solicited machine checks. These are
+	 * created by reset channel path and need not be
+	 * handled here.
+	 */
+	if (crw0->slct) {
+		CIO_CRW_EVENT(2, "solicited machine check for "
+			      "channel path %02X\n", crw0->rsid);
+		return;
+	}
+	chp_id_init(&chpid);
+	chpid.id = crw0->rsid;
+	switch (crw0->erc) {
+	case CRW_ERC_IPARM: /* Path has come. */
+		if (!chp_is_registered(chpid))
+			chp_new(chpid);
+		chsc_chp_online(chpid);
+		break;
+	case CRW_ERC_PERRI: /* Path has gone. */
+	case CRW_ERC_PERRN:
+		chsc_chp_offline(chpid);
+		break;
+	default:
+		CIO_CRW_EVENT(2, "Don't know how to handle erc=%x\n",
+			      crw0->erc);
+	}
+}
+
+int chp_ssd_get_mask(struct chsc_ssd_info *ssd, struct chp_link *link)
+{
+	int i;
+	int mask;
+
+	for (i = 0; i < 8; i++) {
+		mask = 0x80 >> i;
+		if (!(ssd->path_mask & mask))
+			continue;
+		if (!chp_id_is_equal(&ssd->chpid[i], &link->chpid))
+			continue;
+		if ((ssd->fla_valid_mask & mask) &&
+		    ((ssd->fla[i] & link->fla_mask) != link->fla))
+			continue;
+		return mask;
+	}
+	return 0;
+}
+EXPORT_SYMBOL_GPL(chp_ssd_get_mask);
+
+static inline int info_bit_num(struct chp_id id)
+{
+	return id.id + id.cssid * (__MAX_CHPID + 1);
+}
+
+/* Force chp_info refresh on next call to info_validate(). */
+static void info_expire(void)
+{
+	mutex_lock(&info_lock);
+	chp_info_expires = jiffies - 1;
+	mutex_unlock(&info_lock);
+}
+
+/* Ensure that chp_info is up-to-date. */
+static int info_update(void)
+{
+	int rc;
+
+	mutex_lock(&info_lock);
+	rc = 0;
+	if (time_after(jiffies, chp_info_expires)) {
+		/* Data is too old, update. */
+		rc = sclp_chp_read_info(&chp_info);
+		chp_info_expires = jiffies + CHP_INFO_UPDATE_INTERVAL ;
+	}
+	mutex_unlock(&info_lock);
+
+	return rc;
+}
+
+/**
+ * chp_info_get_status - retrieve configure status of a channel-path
+ * @chpid: channel-path ID
+ *
+ * On success, return 0 for standby, 1 for configured, 2 for reserved,
+ * 3 for not recognized. Return negative error code on error.
+ */
+int chp_info_get_status(struct chp_id chpid)
+{
+	int rc;
+	int bit;
+
+	rc = info_update();
+	if (rc)
+		return rc;
+
+	bit = info_bit_num(chpid);
+	mutex_lock(&info_lock);
+	if (!chp_test_bit(chp_info.recognized, bit))
+		rc = CHP_STATUS_NOT_RECOGNIZED;
+	else if (chp_test_bit(chp_info.configured, bit))
+		rc = CHP_STATUS_CONFIGURED;
+	else if (chp_test_bit(chp_info.standby, bit))
+		rc = CHP_STATUS_STANDBY;
+	else
+		rc = CHP_STATUS_RESERVED;
+	mutex_unlock(&info_lock);
+
+	return rc;
+}
+
+/* Return configure task for chpid. */
+static enum cfg_task_t cfg_get_task(struct chp_id chpid)
+{
+	return chp_cfg_task[chpid.cssid][chpid.id];
+}
+
+/* Set configure task for chpid. */
+static void cfg_set_task(struct chp_id chpid, enum cfg_task_t cfg)
+{
+	chp_cfg_task[chpid.cssid][chpid.id] = cfg;
+}
+
+/* Perform one configure/deconfigure request. Reschedule work function until
+ * last request. */
+static void cfg_func(struct work_struct *work)
+{
+	struct chp_id chpid;
+	enum cfg_task_t t;
+	int rc;
+
+	mutex_lock(&cfg_lock);
+	t = cfg_none;
+	chp_id_for_each(&chpid) {
+		t = cfg_get_task(chpid);
+		if (t != cfg_none) {
+			cfg_set_task(chpid, cfg_none);
+			break;
+		}
+	}
+	mutex_unlock(&cfg_lock);
+
+	switch (t) {
+	case cfg_configure:
+		rc = sclp_chp_configure(chpid);
+		if (rc)
+			CIO_MSG_EVENT(2, "chp: sclp_chp_configure(%x.%02x)="
+				      "%d\n", chpid.cssid, chpid.id, rc);
+		else {
+			info_expire();
+			chsc_chp_online(chpid);
+		}
+		break;
+	case cfg_deconfigure:
+		rc = sclp_chp_deconfigure(chpid);
+		if (rc)
+			CIO_MSG_EVENT(2, "chp: sclp_chp_deconfigure(%x.%02x)="
+				      "%d\n", chpid.cssid, chpid.id, rc);
+		else {
+			info_expire();
+			chsc_chp_offline(chpid);
+		}
+		break;
+	case cfg_none:
+		/* Get updated information after last change. */
+		info_update();
+		mutex_lock(&cfg_lock);
+		cfg_busy = 0;
+		mutex_unlock(&cfg_lock);
+		wake_up_interruptible(&cfg_wait_queue);
+		return;
+	}
+	queue_work(chp_wq, &cfg_work);
+}
+
+/**
+ * chp_cfg_schedule - schedule chpid configuration request
+ * @chpid - channel-path ID
+ * @configure - Non-zero for configure, zero for deconfigure
+ *
+ * Schedule a channel-path configuration/deconfiguration request.
+ */
+void chp_cfg_schedule(struct chp_id chpid, int configure)
+{
+	CIO_MSG_EVENT(2, "chp_cfg_sched%x.%02x=%d\n", chpid.cssid, chpid.id,
+		      configure);
+	mutex_lock(&cfg_lock);
+	cfg_set_task(chpid, configure ? cfg_configure : cfg_deconfigure);
+	cfg_busy = 1;
+	mutex_unlock(&cfg_lock);
+	queue_work(chp_wq, &cfg_work);
+}
+
+/**
+ * chp_cfg_cancel_deconfigure - cancel chpid deconfiguration request
+ * @chpid - channel-path ID
+ *
+ * Cancel an active channel-path deconfiguration request if it has not yet
+ * been performed.
+ */
+void chp_cfg_cancel_deconfigure(struct chp_id chpid)
+{
+	CIO_MSG_EVENT(2, "chp_cfg_cancel:%x.%02x\n", chpid.cssid, chpid.id);
+	mutex_lock(&cfg_lock);
+	if (cfg_get_task(chpid) == cfg_deconfigure)
+		cfg_set_task(chpid, cfg_none);
+	mutex_unlock(&cfg_lock);
+}
+
+static int cfg_wait_idle(void)
+{
+	if (wait_event_interruptible(cfg_wait_queue, !cfg_busy))
+		return -ERESTARTSYS;
+	return 0;
+}
+
+static int __init chp_init(void)
+{
+	struct chp_id chpid;
+	int ret;
+
+	ret = crw_register_handler(CRW_RSC_CPATH, chp_process_crw);
+	if (ret)
+		return ret;
+	chp_wq = create_singlethread_workqueue("cio_chp");
+	if (!chp_wq) {
+		crw_unregister_handler(CRW_RSC_CPATH);
+		return -ENOMEM;
+	}
+	INIT_WORK(&cfg_work, cfg_func);
+	init_waitqueue_head(&cfg_wait_queue);
+	if (info_update())
+		return 0;
+	/* Register available channel-paths. */
+	chp_id_for_each(&chpid) {
+		if (chp_info_get_status(chpid) != CHP_STATUS_NOT_RECOGNIZED)
+			chp_new(chpid);
+	}
+
+	return 0;
+}
+
+subsys_initcall(chp_init);

drivers/s390/cio/chp.h

@@ -0,0 +1,72 @@
+/*
+ *    Copyright IBM Corp. 2007, 2010
+ *    Author(s): Peter Oberparleiter <peter.oberparleiter@de.ibm.com>
+ */
+
+#ifndef S390_CHP_H
+#define S390_CHP_H S390_CHP_H
+
+#include <linux/types.h>
+#include <linux/device.h>
+#include <linux/mutex.h>
+#include <asm/chpid.h>
+#include "chsc.h"
+#include "css.h"
+
+#define CHP_STATUS_STANDBY		0
+#define CHP_STATUS_CONFIGURED		1
+#define CHP_STATUS_RESERVED		2
+#define CHP_STATUS_NOT_RECOGNIZED	3
+
+#define CHP_ONLINE 0
+#define CHP_OFFLINE 1
+#define CHP_VARY_ON 2
+#define CHP_VARY_OFF 3
+
+struct chp_link {
+	struct chp_id chpid;
+	u32 fla_mask;
+	u16 fla;
+};
+
+static inline int chp_test_bit(u8 *bitmap, int num)
+{
+	int byte = num >> 3;
+	int mask = 128 >> (num & 7);
+
+	return (bitmap[byte] & mask) ? 1 : 0;
+}
+
+
+struct channel_path {
+	struct device dev;
+	struct chp_id chpid;
+	struct mutex lock; /* Serialize access to below members. */
+	int state;
+	struct channel_path_desc desc;
+	struct channel_path_desc_fmt1 desc_fmt1;
+	/* Channel-measurement related stuff: */
+	int cmg;
+	int shared;
+	void *cmg_chars;
+};
+
+/* Return channel_path struct for given chpid. */
+static inline struct channel_path *chpid_to_chp(struct chp_id chpid)
+{
+	return channel_subsystems[chpid.cssid]->chps[chpid.id];
+}
+
+int chp_get_status(struct chp_id chpid);
+u8 chp_get_sch_opm(struct subchannel *sch);
+int chp_is_registered(struct chp_id chpid);
+struct channel_path_desc *chp_get_chp_desc(struct chp_id chpid);
+void chp_remove_cmg_attr(struct channel_path *chp);
+int chp_add_cmg_attr(struct channel_path *chp);
+int chp_update_desc(struct channel_path *chp);
+int chp_new(struct chp_id chpid);
+void chp_cfg_schedule(struct chp_id chpid, int configure);
+void chp_cfg_cancel_deconfigure(struct chp_id chpid);
+int chp_info_get_status(struct chp_id chpid);
+int chp_ssd_get_mask(struct chsc_ssd_info *, struct chp_link *);
+#endif /* S390_CHP_H */